CN118035387B - A data selection and output method and device based on large language model - Google Patents
A data selection and output method and device based on large language model Download PDFInfo
- Publication number
- CN118035387B CN118035387B CN202410406332.3A CN202410406332A CN118035387B CN 118035387 B CN118035387 B CN 118035387B CN 202410406332 A CN202410406332 A CN 202410406332A CN 118035387 B CN118035387 B CN 118035387B
- Authority
- CN
- China
- Prior art keywords
- request
- selection
- target
- data
- label
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3347—Query execution using vector based model
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3325—Reformulation based on results of preceding query
- G06F16/3326—Reformulation based on results of preceding query using relevance feedback from the user, e.g. relevance feedback on documents, documents sets, document terms or passages
- G06F16/3328—Reformulation based on results of preceding query using relevance feedback from the user, e.g. relevance feedback on documents, documents sets, document terms or passages using graphical result space presentation or visualisation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/338—Presentation of query results
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
本申请公开一种基于大语言模型的数据选取、输出方法和装置，所述选取方法包括：响应于智能服务交互界面输入的数据选取请求，基于大语言模型对所述数据选取请求进行解析，确定所述数据选取请求的选取请求向量；在构建的向量知识库中，基于所述大语言模型查找与所述选取请求向量匹配的目标标签和目标请求示例；根据所述目标标签和所述目标请求示例，基于所述大语言模型生成用于在所述智能服务交互界面输出的与所述数据选取请求对应的选取信息。从而能够在面对大量数据标签的情况下，提高标签匹配、选择等方面的效率，同时也能够避免选择错误而导致的后续对数据利用上的困扰。
The present application discloses a method and device for data selection and output based on a large language model, wherein the selection method includes: in response to a data selection request inputted in an intelligent service interaction interface, parsing the data selection request based on a large language model to determine a selection request vector of the data selection request; in a constructed vector knowledge base, searching for a target label and a target request example matching the selection request vector based on the large language model; based on the target label and the target request example, generating selection information corresponding to the data selection request outputted in the intelligent service interaction interface based on the large language model. This can improve the efficiency of label matching, selection, etc. in the face of a large number of data labels, and can also avoid subsequent troubles in data utilization caused by incorrect selection.
Description
Technical Field
The application relates to the field of artificial intelligence, in particular to a data selection method and device based on a large language model. The application also relates to a data output method and device based on the large language model, a data interaction method and device based on the large language model, a data selection system based on the large language model, a computer storage medium, an electronic device and a computer program product.
Background
Based on the continuous development of the internet and computer technology, the technology related to electronic commerce is updated and iterated continuously, so as to provide efficient services for users with different demands in the whole service link.
In electronic commerce based on internet and computer technology, the selection means that a product provider or a third party carefully selects and organizes commodities according to market demands, user preferences, sales data and other information, so that the types, quality and the like of products can be ensured, and target user groups can be mined. The roles of the options include, but are not limited to: attracting user traffic, improving conversion, optimizing inventory management, etc. Through accurate selection strategy, the application service platform can better satisfy different user's demands to improve the service conversion rate that the product supplied with the user. Therefore, the selection is a key link in meeting the demands of users by improving the conversion rate when on-line and off-line service is performed by depending on an application service platform for the service end for providing commodities.
Disclosure of Invention
The application provides a data selection method based on a large language model, which aims to solve the problems of low data selection efficiency and high selection error rate in the prior art.
The application provides a data selection method based on a large language model, which comprises the following steps:
responding to a data selection request input by an intelligent service interaction interface, analyzing the data selection request based on a large language model, and determining a selection request vector of the data selection request;
Searching a target label and a target request example matched with the selected request vector on the basis of the large language model in the constructed vector knowledge base;
and generating selection information corresponding to the data selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
In some embodiments, the searching, in the constructed vector knowledge base, for the target tag and the target request instance matching the selection request vector based on the large language model includes:
according to text vectorization processing of the history tag information and the history request examples, acquiring history tag vector data and history request example vector data;
constructing a tag vector knowledge base corresponding to the historical tag vector data and a request example vector knowledge base corresponding to the historical request example vector data according to the classification vector index of the storage and query service system;
Searching the target label matched with the selection request vector in the label vector knowledge base based on the large language model according to the matching requirement;
searching the target request examples matched with the selected request vector based on the large language model in the request example vector knowledge base.
In some embodiments, said looking up, in said tag vector knowledge base, said target tag matching said pick request vector based on said large language model according to a matching requirement comprises:
When the matching requirement comprises a label type requirement, the large language model searches a target classification label matched with the selected request vector in the label vector knowledge base according to the classification type of the selected label of the data;
and determining the target classification label as the target label.
In some embodiments, the determining the target classification tag as the target tag includes:
And screening the target classification labels according to label screening conditions, and determining the target classification labels meeting the label screening conditions as the target labels.
In some embodiments, the generating selection information corresponding to the data selection request for output at the intelligent service interaction interface based on the large language model according to the target tag and the target request example includes:
Combining the target classification labels according to the large language model prompt template to determine label combination;
Generating selection information corresponding to the data selection request for output at the intelligent service interactive interface according to the label combination and the assembly of the target request example;
Or alternatively
Combining the target classification labels according to the large language model prompt template to determine label combination;
Checking the label combination, and determining whether a label value corresponding to the target classification label is in a set label value selection range;
if yes, generating selection information corresponding to the data selection request for output on the intelligent service interactive interface according to the label combination and the assembly of the target request example.
In some embodiments, further comprising obtaining at least one of:
Acquiring the type information of a requester of the data selection request;
acquiring output requirement information of the selected information;
The generating selection information corresponding to the data selection request for output at the intelligent service interactive interface according to the label combination and the assembly of the target request example comprises the following steps:
And assembling at least one of the requester type information and the output requirement information with the target tag and the target request example to generate selection information corresponding to the data selection request for output on the intelligent service interactive interface.
In some embodiments, the generating selection information corresponding to the data selection request for output at the intelligent service interaction interface based on the large language model according to the target tag and the target request example includes:
generating the selection information based on the large language model according to the target label and the target request example;
and constructing an operation control of the selection information and/or constructing interactive reference information of the selection information.
The application also provides a data selection output method based on the large language model, which comprises the following steps:
responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information in the data selection method based on the large language model;
Rendering and outputting the selected information on the intelligent service interactive interface;
and responding to the trigger of a data generation engine in the selection information, and outputting the selection data corresponding to the selection information in the intelligent service interaction interface.
The application also provides a data selection interaction method based on the large language model, which comprises the following steps:
responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information in the data selection method based on the large language model;
outputting the selection information to the intelligent service interaction interface;
And responding to a data generation request of the selection information displayed in the intelligent service interaction interface, and outputting the acquired selection data generated according to the selection information to the intelligent service interaction interface.
The application also provides a data selection system based on the large language model, which comprises: the system comprises an interaction end, a processing platform end and a selection engine end;
the interaction end inputs a data selection request through a provided intelligent service interaction interface and sends the data selection request to the processing platform end;
The processing platform end analyzes the data selection request through a large language model and determines a selection request vector of the data selection request; searching a target label and a target request example matched with the selected request vector in a constructed vector knowledge base through the large language model; generating selection information corresponding to the data selection request based on the large language model according to the target label and the target request example, and sending the selection information to the interaction end for output;
the selection engine end selects the selection data corresponding to the selection information through the generation request of the selection information output by the interaction end, and returns the selection data to the interaction end, and the interaction end outputs the selection data on the intelligent interaction interface.
The application also provides a choice selecting method based on the large language model, which comprises the following steps:
Responding to a commodity selection request input by an intelligent service interaction interface, analyzing the commodity selection request based on a large language model, and determining a selection request vector of the commodity selection request;
Searching a target label and a target request example matched with the selected request vector on the basis of the large language model in the constructed vector knowledge base;
And generating selection information corresponding to the commodity selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
The application also provides a computer storage medium for storing the network platform generated data and a program for processing the network platform generated data;
The program, when read and executed by a processor, executes the data selection method based on the large language model; or executing the data selection output method based on the large language model; or executing the data selection interaction method based on the large language model; or perform the option method based on a large language model as described above.
The present application also provides an electronic device including:
A processor;
A memory for storing a program for processing data generated by a network platform, which when read and executed by the processor, performs the above-described large language model-based data selection method; or executing the data selection output method based on the large language model; or executing the data selection interaction method based on the large language model; or perform the option method based on a large language model as described above.
The present application also provides a computer program product comprising a computer program and/or instructions which, when executed by a processor, implement a method of data selection based on a large language model as described above; or executing the data selection output method based on the large language model; or executing the data selection interaction method based on the large language model; or perform the option method based on a large language model as described above.
Compared with the prior art, the application has the following advantages:
according to the data selection method based on the large language model, the vectorization processing is carried out on the input request by utilizing the capacity of the large language model, the matching inquiry of the labels and the examples is carried out according to the vectorization data, the corresponding selection information is obtained according to the matched target labels and the target examples, and the corresponding selection data can be obtained by triggering the generation engine in the selection information, so that the efficiency of label matching, selection and other aspects can be improved under the condition of facing a large number of data labels, and meanwhile, the trouble of subsequent data utilization caused by selection errors can be avoided. In addition, the operation of adding, deleting and modifying the labels in the output selection information can be realized in an interactive mode, and the operation efficiency and experience of the user are improved.
Drawings
FIG. 1 is a flow chart of a data selection method based on a large language model provided by the application.
Fig. 2 is a link diagram of an embodiment of a data selection method based on a large language model provided by the present application.
Fig. 3 is a schematic structural diagram of a data selecting device based on a large language model according to the present application.
FIG. 4 is a flow chart of a data selection output method based on a large language model provided by the application.
Fig. 5 is a schematic diagram of an embodiment of outputting selection information according to the data selection output method based on a large language model provided by the present application.
Fig. 5-1 is a schematic diagram of another embodiment of outputting selection information according to the data selection output method based on a large language model provided by the present application.
Fig. 6 is a schematic diagram of an embodiment of outputting selected data in a data selecting and outputting method based on a large language model according to the present application.
Fig. 7 is a schematic structural diagram of a data selection output device based on a large language model according to the present application.
FIG. 8 is a flow chart of a method for data selection interaction based on a large language model provided by the application.
Fig. 9 is a schematic structural diagram of a data selection interaction device based on a large language model according to the present application.
Fig. 10 is a schematic structural diagram of a data selection system based on a large language model according to the present application.
FIG. 11 is a flow chart of an option method based on a large language model provided by the application.
Fig. 12 is a schematic structural diagram of a selection device based on a large language model according to the present application.
Fig. 13 is a schematic structural diagram of an embodiment of an electronic device provided by the present application.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. The present application may be embodied in many other forms than those herein described, and those skilled in the art will readily appreciate that the present application may be similarly embodied without departing from the spirit or essential characteristics thereof, and therefore the present application is not limited to the specific embodiments disclosed below.
The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The manner of description used in the present application and in the appended claims is for example: "a", "a" and "a" etc. are not limited in number or order, but are used to distinguish the same type of information from each other.
Based on the background art, the application concept of the application is derived from the commodity selection related in the whole application service platform service process, namely the commodity selection in the merchant side or third party operation scene. In the prior art, for selecting articles, manual screening is mainly performed depending on article labels, and the article selecting mode can lead to low operation efficiency under the condition of facing a large amount of articles, and on the other hand, article selecting personnel are required to have label knowledge related to the articles, otherwise, the article selecting error rate is higher.
Based on the problems in the prior art, the application provides a data selection method based on a large language model, and in the embodiment of the method, the implementation process of the method can be understood from the perspective of a service processing platform. Embodiments of the data selection method are described in detail below.
As shown in fig. 1, fig. 1 is a flowchart of a data selection method based on a large language model, where the method includes:
Step S101: responding to a data selection request input by an intelligent service interaction interface, analyzing the data selection request based on a large language model, and determining a selection request vector of the data selection request;
Step S102: searching a target label and a target request example matched with the selected request vector on the basis of the large language model in the constructed vector knowledge base;
Step S103: and generating selection information corresponding to the data selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
Before the above steps S101 to S103 are described in detail, technical words related to the technical solution of the present application are explained.
Large language model: the field of AI artificial intelligence has experienced tremendous growth and change in recent years, and natural language processing (NLP: natural Language Processing) is one of the areas of rapid progress. The large language model (LLM: large Language Model) is used as one of the important components of natural language processing and is mainly used for predicting probability distribution of the next word or character in natural language text, and specifically, the LLM model is a natural language processing model based on deep learning, and can learn grammar and semantics of natural language, so that human-readable text can be generated, and a human-computer interaction function is realized. Human-machine dialogue interaction is implemented in the large language model through Prompt (Prompt), which can be understood as guiding the large language model to generate related natural language text, and the Prompt (Prompt) can be regarded as question information or dialogue information sent to the large language model, or instructions needing to be sent to the large language model, for example: an input sentence of text describes question information, etc.
Intelligent service interaction interface: the service interaction interface is based on an AI intelligent service application (also called AI assistant) of LLM facing to the user, the user's demand information can be input in the interaction interface, and corresponding feedback information can be output aiming at the demand information.
A selection engine: the real-time commodity circling and selecting engine based on label matching can output commodity information to the intelligent service interactive interface in various forms by triggering the commodity circling and selecting engine. The commodity selection scene can be called a selection engine, and the other data selection scenes can be called a data selection engine, so that the engine aims to output corresponding selection data in a plurality of different forms according to a data generation request.
Vector knowledge base: a knowledge base based on vector storage and retrieval is used for quickly inquiring and acquiring information.
In the following, the steps S101 to S103 are described in detail, as shown in fig. 2, and fig. 2 is a link diagram of an embodiment of a data selection method based on a large language model according to the present application.
Regarding step S101: and responding to the data selection request input by the intelligent service interaction interface, analyzing the data selection request based on a large language model, and determining a selection request vector of the data selection request.
The purpose of the step S101 is to obtain semantic information of the data selection request, that is, parse the data selection request and perform vectorization processing, so as to obtain a semantic vector, that is, a selection request vector.
In this embodiment, the parsing of the data selection request may include input questions, user ids, history interaction information, and the like, and the information is vectorized by LLM to obtain a selection request vector representing the semantics, so that the search in step S102 is facilitated.
Regarding step S102: and searching a target label and a target request example matched with the selected request vector on the basis of the large language model in the constructed vector knowledge base.
The purpose of step S102 is to find the target tag and target request instance.
Tags are a data representation that makes data easier to discover, understand, integrate, analyze, and use, while also improving the usability of the data.
The request example may also be understood as a request sample, and the query target request example can improve the matching degree with the data selection request.
The specific implementation process of step S102 may include:
Step S102-1: and acquiring the historical tag vector data and the historical request example vector data according to text vectorization processing of the historical tag information and the historical request examples.
The history tag information may be understood as existing tag information in a selection scene of the selection data, and in this embodiment, the commodity data may be illustrated as an example, that is, the existing tag information of the commodity data. The history request example may be understood as a sample of an existing data selection request. Text vectorization processing of the history tag information and the history request examples is a process of semantic processing through LLM.
Step S102-2: and constructing a tag vector knowledge base corresponding to the historical tag vector data and a request example vector knowledge base corresponding to the historical request example vector data according to the classification vector index of the storage and query service system.
The storage and query service system can provide on-line graph storage and query service, namely storage, query, update and calculation service of large-scale graph data. Of course, the corresponding service can be provided in an offline mode under the condition that the real-time performance requirement is not required or not required, namely, the storage and query service system is not limited to the on-line storage and query service. In this embodiment, a tag vector knowledge base and a request instance vector knowledge base are respectively constructed according to the classification vector indexes of the storage and query service system. The tag vector knowledge base is used for storing tag vectors of historical tag vector data, and the request example vector knowledge base is used for storing request example vectors of history request example vector data.
Step S102-3: and searching the target label matched with the selected request vector based on the large language model in the label vector knowledge base.
Step S102-4: searching the target request examples matched with the selected request vector based on the large language model in the request example vector knowledge base.
The tag vector knowledge base and the request instance vector knowledge base provide a target tag matching the pick request vector and a target request instance matching the pick request vector, respectively.
The specific implementation process of step S102-3 may be that the target tag matched with the selection request vector is searched in the tag vector knowledge base according to the matching requirement based on the large language model. The matching requirements may include tag type requirements, that is, based on the commodity selection scene (such as a meal) provided in the embodiment, tags may be divided into: the type label, the continuous label and the Boolean label are used for enabling the label to be clear and accurate in representation, and further improving the label matching efficiency.
The category type tag may be: cities, brand names, merchant camping categories, merchant category levels, etc., i.e., classifying such data into category labels.
The continuous tag may be: the number of packages, commodity selling price, commodity activity discount and the like can be represented in the form of intervals of [,3], [10, 20], namely, the data are classified as continuous labels.
The boolean tag may be: whether new, staple, entree, sign, quality, etc., i.e., such data is categorized as boolean labels.
When the matching requirement is a tag type requirement, the specific implementation procedure of the step S102-3 may include:
Step S102-31: the large language model searches a target classification label matched with the selection request vector in the label vector knowledge base according to the classification type of the data selection label;
step S102-32: and determining the target classification label as the target label.
In order to improve the matching degree with the input data selection request and reduce the error rate of the subsequent generated selection information, the target classification labels are screened, and labels which do not meet the requirements are removed. Thus, the step S102-32 may specifically include:
Step S102-321: and screening the target classification labels according to label screening conditions, and determining the target classification labels meeting the label screening conditions as the target labels. In this embodiment, the relevant regular matching process may be performed on the target classification label to obtain a label result in json format, and then the label result is screened. The tag screening condition may be a blacklist requirement, namely: removing labels which do not meet the requirements; the mutually exclusive requirements may be that: one of the tags that cannot occur simultaneously is removed, for example: merchant secondary category and merchant primary category, 7 day commodity transaction data and 30 day commodity transaction data. May be a label rule requirement, for example: only the tags in tag sets 1, 2, 3 are needed; the tag in tag set 1 must have its tag value enumerated for that tag; the label value of the label in the label set 2 must be a numerical value, and if a section is output, both end values of the section must be numerical values or the like.
Regarding step S103: and generating selection information corresponding to the data selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
The purpose of step S103 is to generate selection information output at the intelligent service interaction interface.
The specific implementation process can comprise the following steps:
Step S103-11: merging the target classification labels according to the large language model prompt (prompt) template to determine label combinations; the large language model prompt template can be combined with an application scene selected by data to set a structure or format of a prompt so as to guide the content generated by the large language model. And the label combination can be checked to determine whether the label value corresponding to the target classification label is in the set label value selection range. Of course, distance checking of tag values, semantic vector checking, etc. may also be included. Here, the verification of the tag combination may be performed during the process of screening for the tag-compliant screening condition.
Step S103-12: and generating selection information corresponding to the data selection request for output on the intelligent service interactive interface according to the label combination and the assembly of the target request example. The label combination and the target request example can be automatically assembled through the large language model prompt template, and selection information corresponding to the data selection request and used for being output on the intelligent service interaction interface is generated based on the assembled vector information.
In this embodiment, in order to improve the accuracy of selecting information, the tag combination may be checked, as described in the above step S103-12, to determine whether the tag value corresponding to the target classification tag is within the set tag value selection range, so the step S103 generates the selection information corresponding to the data selection request for output at the intelligent service interaction interface, and may be obtained by performing the above step S103-11 and step S103-12, or by adding the step of checking the tag combination in the above step S103-11 and step S103-12, that is: checking the label combination, and determining whether a label value corresponding to the target classification label is in a set label value selection range; if yes, executing the step S103-12.
It may be appreciated that the large language model hint (prompt) template may not only automatically assemble for the tag combination and the target request example, but also may also automatically assemble any one or more of the obtained requester type information of the data selection request, the obtained output requirement information of the selection information, and tag rule information, etc. as parameters of the large language model hint (prompt) template assembly, to generate the selection information according to the tag combination and the target request example, and other parameters.
It should be noted that when the selection information is generated, the method may further include generating prompt (prompt) interaction information corresponding to the selection information and/or the data selection request according to a history interaction memory Id list generated by current interaction of the large language model, and the prompt (prompt) interaction information is used for outputting the prompt (prompt) interaction information at the intelligent service interaction interface.
Further still include: and verifying and matching the target label in the generated selection information according to the label in the data selection engine, so that the data selection engine can identify the target label, and selection data is generated based on the target label. In the option scenario of this embodiment, the data selection engine may be an option engine. The selection engine can acquire corresponding commodities based on the selection information and provide the commodities to the intelligent service interaction interface for output.
In order to improve the interaction experience of the user based on the intelligent service interaction interface, improve the convenience of operation, and facilitate learning the data, the step S103 may further include:
step S103-21: generating the selection information based on the large language model according to the target label and the target request example;
Step S103-22: and constructing an operation control of the selection information and/or constructing interactive reference information of the selection information. The operation control for selecting the information may include: an output data structure built for the target label that can embed an operation control, for example: one or more operation controls such as modification, deletion, etc. are performed on the target label, wherein the modification and deletion can be the operation on the label value of the target label. The tag and tag value may be considered a target tag. And the recommendation label based on the target label and the reference information of the selected data distribution information corresponding to the recommendation label can be constructed. Output data structures of the selection reference information, which may also be constructed based on the selection information, for example: outputting advice information of a data selection mode in the dialogue interaction process; the specific output manner of the recommendation information with association relation and the like constructed for the selection information and used for recommending to the user can refer to fig. 5-1.
The above is a description of the embodiment of the data selection method based on the large language model, which is provided by the application, and can improve the efficiency of label matching, selection and the like under the condition of facing a large number of data labels, and can avoid the trouble of subsequent data utilization caused by error selection.
The foregoing is a specific description of an embodiment of a data selecting method based on a large language model, corresponding to the foregoing embodiment of the data selecting method based on a large language model, and the application also discloses an embodiment of a data selecting device based on a large language model, please refer to fig. 3, and since the embodiment of the device is substantially similar to the embodiment of the method, the description is simpler, and the relevant points refer to part of the description of the embodiment of the method. The device embodiments described below are merely illustrative.
As shown in fig. 3, fig. 3 is a schematic structural diagram of a data selecting device based on a large language model according to the present application, and an embodiment of the device may include: a determining unit 301, a matching unit 302, and a generating unit 303;
The determining unit 301 is configured to respond to a data selection request input by the intelligent service interaction interface, parse the data selection request based on a large language model, and determine a selection request vector of the data selection request; for the specific content of the determination unit 301, reference is made to the description of step S101 described above, and details thereof will not be described here.
The matching unit 302 is configured to search, in the constructed vector knowledge base, a target tag and a target request example that match the selection request vector based on the large language model; the matching unit 302 includes: the system comprises an acquisition subunit, a construction subunit, a first searching subunit and a second searching subunit. The obtaining subunit is configured to obtain, according to text vectorization processing on the history tag information and the history request instance, history tag vector data and history request instance vector data. The construction subunit is configured to construct a tag vector knowledge base corresponding to the historical tag vector data and a request example vector knowledge base corresponding to the historical request example vector data according to the classification vector index of the storage and query service system; the first searching subunit is configured to search, in the tag vector knowledge base, the target tag that matches the selection request vector based on the large language model, and specifically may search, in the tag vector knowledge base according to a matching requirement, the target tag that matches the selection request vector based on the large language model. The second searching subunit is configured to search, in the request instance vector knowledge base, the target request instance that matches the selected request vector based on the large language model. The first searching subunit comprises a target searching subunit and a target determining subunit, and the target searching subunit is used for searching the target classification label matched with the selection request vector in the label vector knowledge base according to the classification type of the data selection label by the large language model when the matching requirement comprises a label type requirement. The target determination subunit determines the target classification tag as the target tag. The target determining subunit is specifically configured to screen the target classification label according to a label screening condition, and determine the target classification label that meets the label screening condition as the target label.
The generating unit 303 is configured to generate, based on the large language model, selection information corresponding to the data selection request, for outputting at the intelligent service interaction interface, according to the target tag and the target request example.
The generating unit 303 includes: the merging subunit and the generating subunit.
The merging subunit is used for merging the target classification labels according to the large language model prompt template to determine label combinations;
and the generation subunit is used for generating selection information corresponding to the data selection request for output on the intelligent service interactive interface according to the label combination and the assembly of the target request example.
Further, the method can also comprise the following steps: and the verification unit is used for verifying the label combination and determining whether the label value corresponding to the target classification label is in the set label value selection range.
Further, the method can also comprise the following steps: the first acquisition unit is used for acquiring the requester type information of the data selection request; and the second acquisition unit is used for acquiring the output requirement information of the selection information. The generating unit is specifically configured to assemble at least one of the requester type information and the output requirement information with the target tag and the target request example, and generate selection information corresponding to the data selection request, where the selection information is used for being output at the intelligent service interaction interface.
Further, the method can also comprise the following steps: and the verification unit is used for verifying and matching the generated target label in the selection information according to the label in the data selection engine.
The generating unit 303 further includes: generating a subunit and constructing a subunit; the generation subunit is configured to generate the selection information based on the large language model according to the target tag and the target request example. The construction subunit is used for constructing the operation control of the selection information and/or constructing the interactive reference information of the selection information.
The foregoing is a schematic description of an embodiment of a data selecting device based on a large language model, and reference may be made to the related content of the foregoing method embodiment for specific content of the embodiment of the device, which is not described in detail herein.
Based on the foregoing, the present application further provides a data selection output method based on a large language model, as shown in fig. 4, fig. 4 is a flowchart of the data selection output method based on the large language model, where an embodiment of the method is described by taking an intelligent service interaction client (which may also be understood as a UI front end) as an example, that is, from the perspective of an intelligent service interaction interface, and specifically includes:
Step S401: responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information generated in a data selection method based on a large language model;
step S402: rendering and outputting the selected information on the intelligent service interactive interface;
Step S403: and responding to the trigger of a data generation engine in the selection information, and outputting the selection data corresponding to the selection information in the intelligent service interaction interface.
Fig. 5, fig. 5-1, and fig. 6 show a schematic diagram of an embodiment of outputting selection information in a data selection output method based on a large language model according to the present application, and fig. 5-1 shows a schematic diagram of another embodiment of outputting selection information in a data selection output method based on a large language model according to the present application; fig. 6 is a schematic diagram of an embodiment of outputting selected data in a data selecting and outputting method based on a large language model according to the present application.
In this embodiment, the purpose of the step S401 is to input a request for selecting data by a user through an intelligent service interaction interface, for example: the input data selection request is "the selected city is Shanghai, the merchant major category is fried rice, and the price is more than 20 dishes". According to the data selection method of the large language model, the input data selection request is analyzed to obtain the vector information of the selection request, and the target label and the target request example matched with the selection request vector are searched in the vector knowledge base, namely: the target labels are city, merchant camping categories and commodity prices, and the label values are respectively as follows: shanghai, fried rice, [0, 20]. And outputting the target labels and the corresponding label values in an intelligent service interactive interface, and outputting a chart corresponding to the target labels. And controls and the like which correspond to the selection information and generate corresponding selection data are arranged in the intelligent service interaction interface. As shown in fig. 6, the selection data may be output in the form of a selection data set, which may be referred to as a selection pool in the scene reference example provided in this embodiment. Detailed information corresponding to the selected data may be output in response to a viewing operation of the selected data set. Of course, the operations that can be performed on the selected data set are not limited to viewing, and may also include operations of deleting, copying, etc., and specific operations may be added or deleted in combination with actual requirements.
The information selected in step S401 may be obtained by referring to the contents of steps S101 to S103, which are not described in detail herein.
It may be appreciated that, for the selection information output in the intelligent service interaction interface, the selection information may include tag information corresponding to the data selection request, and a chart corresponding to a tag value in the tag information, where the form adopted by the chart is not limited to the example given in fig. 6, and may also include a sample generated for the data selection request, so that the selection information output in the intelligent service interaction interface is not limited to the example set forth in the embodiment, and may also include other content related to the selection information, for example: if the target label is not matched, a prompt related to the selection failure can be output.
The above description is provided for an embodiment of a data selection output method based on a large language model, and the embodiment of the method can input a data selection request through an intelligent service interactive interface provided based on the large language model, obtain selection information corresponding to the data selection request, and respond to the operation of a generation engine in the selection information to output selection data corresponding to the selection information, so that when data selection is performed on massive data, the operation is more convenient, the selection efficiency is higher, and convenience and quality are improved for the subsequent use and processing of related data through the selection data.
The foregoing is a specific description of an embodiment of a data selecting and outputting method based on a large language model, corresponding to the foregoing embodiment of a data selecting and outputting device based on a large language model, and the application also discloses an embodiment of a data selecting device based on a large language model, please refer to fig. 7, and since the device embodiment is substantially similar to the method embodiment, the description is simpler, and the relevant points refer to a part of the description of the method embodiment. The device embodiments described below are merely illustrative.
As shown in fig. 7, fig. 7 is a schematic structural diagram of a data selection output device based on a large language model, where an embodiment of the device may include:
an obtaining unit 701, configured to obtain selection information in the data selection method based on the large language model in response to a data selection request input by the intelligent service interaction interface;
the first output unit 702 is configured to render and output the selected information to the intelligent service interaction interface;
and a second output unit 703, configured to output, in response to a trigger to the data generation engine in the selection information, selection data corresponding to the selection information in the intelligent service interaction interface.
For the specific content of the embodiment of the apparatus, reference may be made to the content of the above steps S101 to S103 and the above steps S401 to S403, which will not be described in detail.
Based on the above-mentioned application, a data selection interaction method based on a large language model is also provided, as shown in fig. 8, fig. 8 is a flowchart of a data selection interaction method based on a large language model, which includes:
Step S801: responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information generated according to a data selection method based on a large language model;
Step S802: outputting the selection information to the intelligent service interaction interface;
Step S803: and responding to a data generation request of the selection information displayed in the intelligent service interaction interface, and outputting the acquired selection data generated according to the selection information to the intelligent service interaction interface.
The description of the steps S801 to S803 is mainly that of the request input angle, the large model processing angle, the selected data generating angle, and the like, and specific contents may refer to the above steps S101 to S103, and the related contents of the steps S401 to S403, which will not be described in detail herein.
Correspondingly, the application also provides a data selection interaction device based on a large language model, as shown in fig. 9, the embodiment of the device can include:
an obtaining unit 901, configured to obtain selection information generated according to a data selection method based on a large language model in response to a data selection request input by an intelligent service interaction interface;
a first output unit 902, configured to output the selection information to the intelligent service interaction interface;
and the second output unit 903 is configured to output, in response to a data generation request for the selection information displayed in the intelligent service interaction interface, the obtained selection data generated according to the selection information to the intelligent service interaction interface.
For the specific content of the apparatus, reference may be made to the above-mentioned steps S101 to S103 and the related content of steps S401 to S403, which will not be described in detail here.
Based on the foregoing, the present application further provides a data selection system based on a large language model, as shown in fig. 10, where an embodiment of the system may include: an interaction end 1001, a processing platform end 1002 and a selection engine end 1003;
The interactive terminal 1001 inputs a data selection request through the provided intelligent service interactive interface, and sends the data selection request to the processing platform terminal.
The processing platform 1002 analyzes the data selection request through a large language model, and determines a selection request vector of the data selection request; searching a target label and a target request example matched with the selected request vector in a constructed vector knowledge base through the large language model; and generating selection information corresponding to the data selection request based on the large language model according to the target label and the target request example, and sending the selection information to the interaction end for output. The processing platform end may be an online service capability with SERVERLESS (no server computing) for searching, recommending, etc. services.
The selection engine 1003 selects the selection data corresponding to the selection information according to the generation request of the selection information output by the interaction end, and returns the selection data to the interaction end, where the interaction end outputs the selection data on the intelligent interaction interface.
Based on the foregoing, the present application further provides a selection method based on a large language model, as shown in fig. 11, where an embodiment of the selection method may include:
step S1101: responding to a commodity selection request input by an intelligent service interaction interface, analyzing the commodity selection request based on a large language model, and determining a selection request vector of the commodity selection request;
step S1102: searching a target label and a target request example matched with the selected request vector on the basis of the large language model in the constructed vector knowledge base;
Step S1103: and generating selection information corresponding to the commodity selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
Further, the method can also comprise the following steps: triggering the generation engine of the item selection information in the intelligent service interaction interface, and outputting an item selection pool corresponding to the item selection information in the intelligent service interaction interface.
Further, the method can also comprise the following steps: and triggering the option pool checking control can output commodity information in the option pool.
Of course, it can be understood that the triggering of the generating engine may be the option pool or the selected commodity information, so that the output display of the option information is not limited to the above, and the requirement of the requester can be met, and the requester can conveniently view and acquire the option information. In this embodiment, only the option pool is illustrated as an implementation.
For details of this alternative method embodiment, reference may be made to the above-mentioned steps S101 to S103, steps S401 to S403, and steps S801 to S803, which will not be described in detail herein.
Correspondingly, the application also provides a choice device based on a large language model, as shown in fig. 12, the embodiment of the device can include:
the determination unit 1201: the method comprises the steps of responding to a commodity selection request input by an intelligent service interaction interface, analyzing the commodity selection request based on a large language model, and determining a selection request vector of the commodity selection request;
Matching unit 1202: the target label and the target request example matched with the selected request vector are searched in the constructed vector knowledge base based on the large language model;
The generation unit 1203: and generating option information corresponding to the commodity selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
Based on the above, the present application also provides a computer storage medium for storing network platform generated data and a program for processing the network platform generated data;
The program, when read and executed by a processor, performs the steps of the large language model-based data selection method described above; or executing the data selection output method based on the large language model; or executing the data selection interaction method based on the large language model; or performing the steps of the option method based on the large language model as described above.
Based on the above, the present application also provides an electronic device, as shown in fig. 13, including:
A processor 1301;
A memory 1302 for storing a program for processing data generated by a network platform, which when read and executed by the processor, performs the steps of the data selection method based on the large language model as described above; or executing the data selection output method based on the large language model; or executing the data selection interaction method based on the large language model; or performing the steps of the option method based on the large language model as described above.
Based on the foregoing, the present application also provides a computer program product, including a computer program and/or instructions which, when executed by a processor, implement the steps of the data selection method based on a large language model as described above; or realizing the data selection output method based on the large language model; or realizing the data selection interaction method based on the large language model; or the steps of implementing the above-described large language model-based option method.
It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are information and data authorized by the user or fully authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards of the related country and region, and provide corresponding operation entries for the user to select authorization or rejection.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
1. Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer readable media, as defined herein, does not include non-transitory computer readable media (transmission media), such as modulated data signals and carrier waves.
2. It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
While the application has been described in terms of preferred embodiments, it is not intended to be limiting, but rather, it will be apparent to those skilled in the art that various changes and modifications can be made herein without departing from the spirit and scope of the application as defined by the appended claims.
Claims (13)
1. A method for selecting data based on a large language model, comprising:
Responding to a data selection request input by an intelligent service interaction interface, analyzing the data selection request based on a large language model, and determining the acquired semantic information of the data selection request as a selection request vector of the data selection request;
According to the matching requirement, searching the target label matched with the selected request vector based on the large language model in the constructed label vector knowledge base, and comprising the following steps: when the matching requirement comprises a label type requirement, the large language model searches a target classification label matched with the selection request vector in the label vector knowledge base according to the classification type of the data selection label, and determines the target classification label as the target label; searching a target request example matched with the selected request vector based on the large language model in the constructed request example vector knowledge base;
and generating selection information corresponding to the data selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
2. The method according to claim 1, wherein the target tag matched with the selected request vector is searched in the constructed tag vector knowledge base based on the large language model according to the matching requirement; searching a target request example matched with the selected request vector based on the large language model in the constructed request example vector knowledge base, wherein the method comprises the following steps of:
according to text vectorization processing of the history tag information and the history request examples, acquiring history tag vector data and history request example vector data;
constructing a tag vector knowledge base corresponding to the historical tag vector data and a request example vector knowledge base corresponding to the historical request example vector data according to the classification vector index of the storage and query service system;
Searching the target label matched with the selection request vector in the label vector knowledge base based on the large language model according to the matching requirement;
searching the target request examples matched with the selected request vector based on the large language model in the request example vector knowledge base.
3. The method of claim 1, wherein the determining the target classification tag as the target tag comprises:
And screening the target classification labels according to label screening conditions, and determining the target classification labels meeting the label screening conditions as the target labels.
4. The method of claim 1, wherein generating selection information corresponding to the data selection request for output at the intelligent service interaction interface based on the large language model according to the target tag and the target request instance comprises:
Combining the target classification labels according to the large language model prompt template to determine label combination;
Generating selection information corresponding to the data selection request for output at the intelligent service interactive interface according to the label combination and the assembly of the target request example;
Or alternatively
Combining the target classification labels according to the large language model prompt template to determine label combination;
Checking the label combination, and determining whether a label value corresponding to the target classification label is in a set label value selection range;
if yes, generating selection information corresponding to the data selection request for output on the intelligent service interactive interface according to the label combination and the assembly of the target request example.
5. The method of claim 1, further comprising obtaining at least one of the following information:
Acquiring the type information of a requester of the data selection request;
acquiring output requirement information of the selected information;
The generating selection information corresponding to the data selection request for output at the intelligent service interactive interface according to the label combination and the assembly of the target request example comprises the following steps:
And assembling at least one of the requester type information and the output requirement information with the target tag and the target request example to generate selection information corresponding to the data selection request for output on the intelligent service interactive interface.
6. The method of claim 1, wherein generating selection information corresponding to the data selection request for output at the intelligent service interaction interface based on the large language model according to the target tag and the target request instance comprises:
generating the selection information based on the large language model according to the target label and the target request example;
and constructing an operation control of the selection information and/or constructing interactive reference information of the selection information.
7. The data selection output method based on the large language model is characterized by comprising the following steps:
Responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information in the method of any one of claims 1 to 6;
Rendering and outputting the selected information on the intelligent service interactive interface;
and responding to the trigger of a data generation engine in the selection information, and outputting the selection data corresponding to the selection information in the intelligent service interaction interface.
8. The data selection interaction method based on the large language model is characterized by comprising the following steps of:
Responding to a data selection request input by an intelligent service interaction interface, and acquiring selection information in the method of any one of claims 1 to 6;
outputting the selection information to the intelligent service interaction interface;
And responding to a data generation request of the selection information displayed in the intelligent service interaction interface, and outputting the acquired selection data generated according to the selection information to the intelligent service interaction interface.
9. A large language model based data selection system, comprising: the system comprises an interaction end, a processing platform end and a selection engine end;
the interaction end inputs a data selection request through a provided intelligent service interaction interface and sends the data selection request to the processing platform end;
The processing platform end analyzes the data selection request through a large language model and determines a selection request vector of the data selection request; according to the matching requirement, searching the target label matched with the selected request vector in the constructed label vector knowledge base through the large language model, wherein the method comprises the following steps: when the matching requirement comprises a label type requirement, the large language model searches a target classification label matched with the selection request vector in the label vector knowledge base according to the classification type of the data selection label, and determines the target classification label as the target label; searching a target request example matched with the selected request vector based on the large language model in the constructed request example vector knowledge base; generating selection information corresponding to the data selection request based on the large language model according to the target label and the target request example, and sending the selection information to the interaction end for output;
The selection engine end selects the selection data corresponding to the selection information through the generation request of the selection information output by the interaction end, and returns the selection data to the interaction end, and the interaction end outputs the selection data on an intelligent interaction interface.
10. A method of selecting items based on a large language model, comprising:
Responding to a commodity selection request input by an intelligent service interaction interface, analyzing the commodity selection request based on a large language model, and determining a selection request vector of the commodity selection request;
According to the matching requirement, searching the target label matched with the selected request vector based on the large language model in the constructed label vector knowledge base, and comprising the following steps: when the matching requirement comprises a label type requirement, the large language model searches a target classification label matched with the selection request vector in the label vector knowledge base according to the classification type of the data selection label, and determines the target classification label as the target label; searching a target request example matched with the selected request vector based on the large language model in the constructed request example vector knowledge base;
And generating selection information corresponding to the commodity selection request for output on the intelligent service interactive interface based on the large language model according to the target label and the target request example.
11. A computer storage medium for storing network platform generated data and a program for processing the network platform generated data;
Which when read by a processor performs the method according to any of claims 1-6; or performing the method of claim 7; or performing the method of claim 8; or perform the method of claim 10.
12. An electronic device, comprising:
A processor;
A memory for storing a program for processing network platform generated data, which program, when read for execution by the processor, performs the method of any one of claims 1-6; or performing the method of claim 7; or performing the method of claim 8; or perform the method of claim 10.
13. A computer program product comprising a computer program and/or instructions which, when executed by a processor, implement the method of any one of claims 1-6; or to implement the method of claim 7; or to implement the method of claim 8; or to implement the method of claim 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410406332.3A CN118035387B (en) | 2024-04-03 | 2024-04-03 | A data selection and output method and device based on large language model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410406332.3A CN118035387B (en) | 2024-04-03 | 2024-04-03 | A data selection and output method and device based on large language model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118035387A CN118035387A (en) | 2024-05-14 |
| CN118035387B true CN118035387B (en) | 2024-09-17 |
Family
ID=90991547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410406332.3A Active CN118035387B (en) | 2024-04-03 | 2024-04-03 | A data selection and output method and device based on large language model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118035387B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118710344A (en) * | 2024-05-15 | 2024-09-27 | 杭州阿里巴巴海外互联网产业有限公司 | Information processing method, device and system |
| CN118839769B (en) * | 2024-06-27 | 2025-08-29 | 北京字跳网络技术有限公司 | Task processing method, device, equipment and computer-readable storage medium |
| CN119358657B (en) * | 2024-12-26 | 2025-03-28 | 中科云谷科技有限公司 | Method, device and storage medium for selecting knowledge base |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117648420A (en) * | 2023-11-30 | 2024-03-05 | 在线途游（北京）科技有限公司 | Intelligent customer service question-answering method and system based on large language model |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107436922B (en) * | 2017-07-05 | 2021-06-08 | 北京百度网讯科技有限公司 | Text label generation method and device |
| KR102800045B1 (en) * | 2022-03-02 | 2025-04-23 | 네이버 주식회사 | Method and system for goods recommendation |
| CN116861087B (en) * | 2023-07-06 | 2025-11-28 | 广州探迹科技有限公司 | Customer intelligent recommendation method, system, equipment and medium based on large language model |
| CN116992027A (en) * | 2023-07-18 | 2023-11-03 | 达而观数据(成都)有限公司 | Text multi-label classification based on generated language model fine tuning |
| CN116911312B (en) * | 2023-09-12 | 2024-01-05 | 深圳须弥云图空间科技有限公司 | Task type dialogue system and implementation method thereof |
| CN117709357A (en) * | 2023-11-09 | 2024-03-15 | 淘宝（中国）软件有限公司 | AI intelligent robot interaction processing method and device based on large language model |
| CN117633219A (en) * | 2023-11-10 | 2024-03-01 | 共道网络科技有限公司 | Complaint classification method, device, equipment and storage medium |
| CN117273868A (en) * | 2023-11-20 | 2023-12-22 | 浙江口碑网络技术有限公司 | Shop recommendation method and device, electronic equipment and storage medium |
| US11947923B1 (en) * | 2023-11-27 | 2024-04-02 | Google Llc | Multimedia content management for large language model(s) and/or other generative model(s) |
| CN117541355A (en) * | 2023-11-30 | 2024-02-09 | 携程计算机技术（上海）有限公司 | Commodity information display method and system |
| CN117350825B (en) * | 2023-12-05 | 2024-03-15 | 浙江口碑网络技术有限公司 | Shopping guide method and system based on large language model |
| CN117633356A (en) * | 2023-12-05 | 2024-03-01 | 百度时代网络技术（北京）有限公司 | Resource processing method and device based on large language model, electronic equipment and medium |
-
2024
- 2024-04-03 CN CN202410406332.3A patent/CN118035387B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117648420A (en) * | 2023-11-30 | 2024-03-05 | 在线途游（北京）科技有限公司 | Intelligent customer service question-answering method and system based on large language model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118035387A (en) | 2024-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118035387B (en) | A data selection and output method and device based on large language model | |
| Felfernig et al. | Constraint-based recommender systems | |
| US20170337261A1 (en) | Decision Making and Planning/Prediction System for Human Intention Resolution | |
| KR20180022762A (en) | Method, system and computer-readable recording medium for providing customer counseling service using real-time response message generation | |
| KR20170137419A (en) | Method, system and computer-readable recording medium for providing customer counseling service using real-time response message generation | |
| CN118277588B (en) | Query request processing method, electronic device and storage medium | |
| CN113781149B (en) | Information recommendation method, device, computer-readable storage medium, and electronic device | |
| CN111090739A (en) | Information processing method, information processing device, electronic device, and storage medium | |
| CN118733903A (en) | Commodity object search method, device, electronic device and storage medium | |
| CN119128277A (en) | A product recommendation method, device and medium based on intelligent agent | |
| US7398227B2 (en) | Methods, systems, and computer for managing purchasing data | |
| CN113377640B (en) | Method, medium, device and computing equipment for explaining model under business scene | |
| CN118861241A (en) | Dialogue generation method, device, computer equipment and storage medium | |
| CN112148939A (en) | Data processing method and device and electronic equipment | |
| KR20180049277A (en) | Method and apparatus for recommendation of financial instruments using chatting user platform | |
| US12450535B2 (en) | Multi-layer micro model analytics framework in information processing system | |
| Han et al. | Developing smart service concepts: morphological analysis using a Novelty-Quality map | |
| CN114429384B (en) | Intelligent product recommendation method and system based on e-commerce platform | |
| CN119719315A (en) | Intelligent customer service session processing method, device and program product | |
| US20240296376A1 (en) | Requirements driven machine learning models for technical configuration | |
| Patel et al. | Personalized context-oriented job recommendation system based on knowledge graph | |
| CN120336133B (en) | Intelligent operation and maintenance method, equipment and storage medium | |
| CN118245529B (en) | User behavior data analysis method and system based on digital person | |
| CN120804275B (en) | Commodity information global association method and system in multi-round dialogue process | |
| US20240296375A1 (en) | Requirements driven machine learning models for technical configuration |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20251212 Address after: Room 507, 788 Zhenbei Road, Putuo District, Shanghai, 200333 Patentee after: Lazas Network Technology (Shanghai) Co., Ltd. Country or region after: China Address before: Room 203, 2nd Floor, Building 5, No. 699, Wangshang Road, Changhe Street, Binjiang District, Hangzhou City, Zhejiang Province, 310056 Patentee before: ZHEJIANG KOUBEI NETWORK TECHNOLOGY Co.,Ltd. Country or region before: China Patentee before: Lazas Network Technology (Shanghai) Co., Ltd. |
|
| TR01 | Transfer of patent right |