CN118069822A - Recommendation method, recommendation device, recommendation equipment and storage medium - Google Patents
Recommendation method, recommendation device, recommendation equipment and storage medium Download PDFInfo
- Publication number
- CN118069822A CN118069822A CN202410108066.6A CN202410108066A CN118069822A CN 118069822 A CN118069822 A CN 118069822A CN 202410108066 A CN202410108066 A CN 202410108066A CN 118069822 A CN118069822 A CN 118069822A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- probability
- data
- tag
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 74
- 238000002372 labelling Methods 0.000 claims description 42
- 238000012549 training Methods 0.000 claims description 31
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 15
- 238000012163 sequencing technique Methods 0.000 claims description 15
- 230000011218 segmentation Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 abstract description 18
- 238000004422 calculation algorithm Methods 0.000 abstract description 13
- 238000010586 diagram Methods 0.000 description 14
- 238000012545 processing Methods 0.000 description 14
- 230000006870 function Effects 0.000 description 13
- 230000001105 regulatory effect Effects 0.000 description 9
- 238000010801 machine learning Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 238000000605 extraction Methods 0.000 description 6
- 238000011160 research Methods 0.000 description 6
- 238000010845 search algorithm Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 3
- 238000007635 classification algorithm Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001151 other effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06F16/3346—Query execution using probabilistic model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention provides a recommendation method, a recommendation device, electronic equipment and a storage medium, wherein the recommendation method comprises the steps of obtaining search requirements of a user; generating a search label and a corresponding label probability; acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability; then matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data; if the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended; and recommending knowledge to the user based on the knowledge to be recommended. By the method, the problem that the keyword retrieval algorithm in the prior art can only match accurate keywords generally and cannot process semantic relativity of synonyms, paraphraseology and the like can be solved, and therefore retrieval accuracy is improved.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a recommendation method, apparatus, device, and storage medium.
Background
In state-run assets regulatory domain, there are currently some big data based regulatory systems. These systems typically provide relevant information by collecting and collating a large amount state-run assets of the relevant data and then utilizing keyword retrieval, which typically uses conventional databases to store the data and search algorithms based on keyword matching to find the relevant information.
However, conventional keyword matching-based search algorithms have some limitations. They can only retrieve according to keywords provided by the user and cannot accurately understand the intention and the demand of the user. Therefore, the accuracy of the search is not high.
Disclosure of Invention
In view of the foregoing, embodiments of the present invention are directed to providing a recommendation method, apparatus, device, and storage medium that overcome, or at least partially solve, the foregoing.
In order to solve the above problems, an embodiment of the present invention discloses a recommendation method, which includes:
Acquiring the retrieval requirement of a user;
Generating a search label and a corresponding label probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag;
Acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability;
Matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data;
If the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended;
and recommending knowledge to the user based on the knowledge to be recommended.
Optionally, the recommending the knowledge to the user based on the knowledge to be recommended includes:
Dividing the search label into a plurality of label sets according to the label probability of the search label and a preset segmentation threshold value;
combining the knowledge to be recommended corresponding to the retrieval tag in the same tag set to obtain a knowledge set to be recommended;
And recommending knowledge to the user based on the knowledge set to be recommended.
Optionally, the recommending knowledge to the user based on the knowledge set to be recommended includes:
Sequencing the knowledge sets to be recommended according to the segmentation sequence corresponding to the tag set to obtain a sequencing result of the knowledge sets to be recommended;
And recommending knowledge to the user according to the sequencing result of the knowledge set to be recommended.
Optionally, the probability knowledge model is obtained through training in the following manner:
Acquiring service data and a type label corresponding to the service data;
Acquiring business knowledge and business rules;
Extracting characteristic knowledge data according to the service knowledge, the service rule, the service data and the type label corresponding to the service data to obtain the characteristic knowledge data and the type label corresponding to the characteristic knowledge data;
Acquiring initial probability aiming at the characteristic knowledge data;
training a preset probability knowledge model based on the feature knowledge data, the type label corresponding to the feature knowledge data and the initial probability to obtain a probability knowledge model.
Optionally, the service data and the type tag corresponding to the service data include:
acquiring service data;
And identifying the service data based on a preset labeling model, determining the type of the service data and generating a type label.
Optionally, the preset labeling model is obtained through training in the following manner:
acquiring historical service data, wherein the historical service data comprises historical service knowledge and historical service rules;
And training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model.
Optionally, the generating, according to the search requirement, a search label and a corresponding label probability includes:
Acquiring a preset demand conversion model;
And inputting the search requirement to the requirement conversion model to obtain the search label and the corresponding label probability.
Correspondingly, the embodiment of the invention discloses a recommending device, which comprises:
the demand acquisition module is used for acquiring the retrieval demand of the user;
The label generating module is used for generating a search label and a corresponding label probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag;
The knowledge generation module is used for acquiring a probability knowledge model, inputting the retrieval tag to the probability knowledge model and obtaining characteristic knowledge data and corresponding recommendation probability;
the probability comparison module is used for matching the retrieval tag with the characteristic knowledge data and comparing the tag probability of the retrieval tag with the recommendation probability of the matched characteristic knowledge data;
the determining module is used for taking the characteristic knowledge data as knowledge to be recommended if the label probability of the search label is larger than the recommendation probability of the characteristic knowledge data;
and the recommending module is used for recommending knowledge to the user based on the knowledge to be recommended.
Optionally, the recommendation module includes:
The sub-splitting module is used for dividing the search tag into a plurality of tag sets according to the tag probability of the search tag and a preset segmentation threshold value;
the combination sub-module is used for combining the knowledge to be recommended corresponding to the retrieval tag in the same tag set to obtain a knowledge set to be recommended;
And the recommending sub-module is used for recommending knowledge to the user based on the knowledge set to be recommended.
Optionally, the recommendation sub-module includes:
The sorting unit is used for sorting the knowledge sets to be recommended according to the segmentation sequence corresponding to the tag set to obtain a sorting result of the knowledge sets to be recommended;
And the recommending unit is used for recommending knowledge to the user according to the sequencing result of the knowledge set to be recommended.
Optionally, the knowledge generation module includes:
The first acquisition sub-module is used for acquiring service data and type labels corresponding to the service data;
the second acquisition sub-module is used for acquiring service knowledge and service rules;
The characteristic knowledge determining submodule is used for extracting characteristic knowledge data according to the service knowledge, the service rule, the service data and the type label corresponding to the service data to obtain the characteristic knowledge data and the type label corresponding to the characteristic knowledge data;
A third obtaining sub-module, configured to obtain an initial probability for the feature knowledge data;
And the training sub-module is used for training a preset probability knowledge model based on the characteristic knowledge data, the type label corresponding to the characteristic knowledge data and the initial probability to obtain a probability knowledge model.
Optionally, the first obtaining sub-module includes:
The acquisition unit is used for acquiring service data;
And the generating unit is used for identifying the service data based on a preset labeling model, determining the type of the service data and generating a type label.
Optionally, the generating unit includes:
the acquisition subunit is used for acquiring historical service data, wherein the historical service data comprises historical service knowledge and historical service rules;
And the training subunit is used for training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model.
Optionally, the tag generation module includes:
the demand conversion model acquisition sub-module is used for acquiring a preset demand conversion model;
and the search label generation sub-module is used for inputting the search requirement to the requirement conversion model to obtain the search label and the corresponding label probability.
Correspondingly, the embodiment of the invention discloses an electronic device, which comprises: a processor, a memory and a computer program stored on the memory and capable of running on the processor, which when executed by the processor carries out the steps of the preferred method embodiments described above.
Accordingly, embodiments of the present invention disclose a computer readable storage medium having a computer program stored thereon, which when executed by a processor, implements the steps of the preferred method embodiments described above.
The embodiment of the invention has the following advantages: the recommendation method provided by the embodiment of the invention needs to acquire the retrieval requirement of the user; then according to the search requirement, generating a search label and a corresponding label probability; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag; acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability; then matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data; if the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended; and recommending knowledge to the user based on the knowledge to be recommended. By the method, the search labels are generated according to the search requirements of the users and are input into the probability knowledge model, so that the problem that the keyword search algorithm in the prior art can only match accurate keywords and cannot process semantic relativity of synonyms, paraphraseology and the like can be solved, and the search accuracy is improved.
Drawings
FIG. 1 is a flow chart of the steps of a recommendation method in an embodiment of the present invention;
FIG. 2 is a schematic diagram of a recommendation method in an embodiment of the invention;
FIG. 3 is a block diagram of a recommendation device in an embodiment of the present invention;
fig. 4 is a block diagram of an electronic device according to an embodiment of the present invention;
fig. 5 is a block diagram of a computer-readable storage medium according to an embodiment of the present invention.
Detailed Description
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
One of the core ideas of the embodiment of the invention is that the recommendation method provided by the embodiment of the invention needs to acquire the retrieval requirement of the user; then according to the search requirement, generating a search label and a corresponding label probability; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag; acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability; then matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data; if the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended; and recommending knowledge to the user based on the knowledge to be recommended. By the method, the search labels are generated according to the search requirements of the users and are input into the probability knowledge model, so that the problem that the keyword search algorithm in the prior art can only match accurate keywords and cannot process semantic relativity of synonyms, paraphraseology and the like can be solved, and the search accuracy is improved.
In practical application, state-run assets has the following technical problems in the supervision field:
Information retrieval accuracy problems. In the prior art, the keyword retrieval algorithm can only be matched with accurate keywords, and cannot process semantic relevance problems such as synonyms, near-meaning words and the like, so that the accuracy of retrieval results is low. The conventional keyword search algorithm mainly searches documents or data containing keywords from a database by matching the keywords input by a user. However, in state-run assets regulatory domains, there are often a large number of synonyms, paraphraseology, and industry-specific terms and abbreviations, with semantic relatedness between the terms. The traditional keyword retrieval algorithm cannot understand the semantic relevance, and can only simply retrieve according to the accurate matching of keywords, so that the accuracy of a retrieval result is low. For example, in state-run assets regulatory domain, a user may search for the keyword "business profit," but documents or data in the database may use the synonym "business profit" or the paraphrasing "business profit. The traditional keyword retrieval algorithm cannot match the related words, so that related documents or data cannot be retrieved, and the accuracy of a retrieval result is affected.
The information recommends personalized questions. The recommendation algorithm in the prior art generally only carries out recommendation based on the historical behaviors and preferences of the user, and cannot fully consider the real-time requirements and personalized characteristics of the user, and the recommendation algorithm in the prior art cannot fully consider the personalized characteristics of the user. Each user has unique interests, preferences and favorites, but the conventional algorithm mainly recommends based on group behaviors and preferences, and cannot fully consider the personalized characteristics of each user, so that the accuracy of recommendation results and user experience are poor. Because the real-time requirements and individuation characteristics of the users cannot be fully considered, the recommendation result of the recommendation algorithm in the prior art is low in accuracy. The recommended results may not match the actual needs of the user, resulting in reduced satisfaction of the user with the recommended results. Poor accuracy of the recommendation results can affect the user experience. If the user frequently receives the recommendation which does not accord with the self demand, the user can be distrust on a recommendation system, and the acceptance and satisfaction of the user on the recommendation result are reduced.
Data processing efficiency problems. In the prior art, a traditional database storage mode is generally used for data processing, so that large-scale data cannot be processed efficiently, and the data processing efficiency is low.
The keyword retrieval algorithm in the prior art cannot process the problem of semantic relativity, so that the accuracy of the retrieval result is low. The recommendation algorithm cannot fully consider the real-time requirements and individuation characteristics of the user, so that the accuracy and the user experience of the recommendation result are poor. In addition, the data processing method in the prior art cannot efficiently process large-scale data, so that the data processing efficiency is low.
Referring to fig. 1, a flowchart illustrating steps of a recommendation method in an embodiment of the present invention may specifically include the following steps:
step 101, obtaining the retrieval requirement of the user.
Specifically, a search requirement of a user is obtained, and the user search requirement refers to description or expression of required information by the user, and the user can express the required information by means of keywords, query conditions, questions and the like.
The embodiment of the invention can be applied to the fields of state-run assets supervision departments, financial institutions, investment institutions, research institutions, academia and the like, and the recommendation method of the embodiment of the invention can enable a user to quickly acquire information related to national asset supervision, including enterprise information, financial data, market data and the like. The information can provide decision support and business guidance for users, help them to perform risk assessment, asset management, investment decision and research analysis, and improve the working efficiency and decision accuracy.
State-run assets the supervision department can utilize the intelligent recommendation and intelligent retrieval functions provided by the embodiment of the invention to quickly acquire and analyze state-run assets supervision related big data, thereby providing scientific and accurate support for decision making. The financial institutions can utilize the intelligent recommendation and intelligent retrieval functions provided by the embodiment of the invention to perform risk assessment and credit rating on national enterprises. The financial institutions are helped to evaluate the risk level of enterprises more accurately through intelligent recommendation of related supervision indexes and risk early warning information, so that risks are managed better and loan policies are formulated. Enterprises can acquire state-run assets relevant market data and industry indexes by utilizing the intelligent recommendation and intelligent retrieval functions provided by the embodiment of the invention, scientific and accurate support is provided for business decisions of the enterprises, and the intelligent recommendation of proper business strategies and market opportunities helps the enterprises optimize the business decisions and improves the competitiveness. Research institutions, academia, consultation companies and the like can utilize the intelligent recommendation and intelligent retrieval functions provided by the embodiment of the invention to conduct deep research and analysis on state-run assets supervision fields. The relevant supervision indexes and rules are recommended intelligently, so that research institutions and academia are helped to better understand and explain state-run assets the supervision rules and trends, and references are provided for policy making and decision making.
Step 102, generating a search label and a corresponding label probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag.
Specifically, after obtaining the search requirement of the user, extracting keywords according to the search requirement of the user, wherein the keywords can be words input by the user in a search box or can be important words extracted from problem description, and intelligently associating the extracted keywords according to the search requirement of the user as labels, classifying the labels, and generating search labels and corresponding label probabilities. The search requirements of the user may be classified according to the search requirements of the user using a classification algorithm or a machine learning model, and corresponding classification labels are generated, from which attributes are extracted as labels using attribute extraction techniques, and the attributes may be specific attributes mentioned by the user in the description of the problem, such as time, place, price, etc., for example. And, the generated search tags and corresponding tag probabilities need to be converted into a form that can be used to intelligently recommend knowledge to the user. There may be a plurality of generated search tags and corresponding tag probabilities, and the generated search tags and corresponding tag probabilities may be [ "funds", 5/8] and [ "profit", 3/8] for example.
And step 103, acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability.
Specifically, a probabilistic knowledge model is obtained, and the probabilistic knowledge model can output feature knowledge data, a tag corresponding to the feature knowledge data, and a recommendation probability corresponding to the feature knowledge data. And inputting the retrieval labels obtained according to the retrieval requirements of the users into the probability knowledge model, so that a plurality of pieces of characteristic knowledge data, labels corresponding to the characteristic knowledge data and recommendation probabilities corresponding to the characteristic knowledge data can be obtained. For example, the feature knowledge data, the tags corresponding to the feature knowledge data, and the recommendation probability corresponding to the feature knowledge data may be [ "large enterprise funds approval", "funds", 5/8].
And 104, matching the retrieval tag with the characteristic knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched characteristic knowledge data.
Specifically, correlation matching is performed on the retrieval tag and the feature knowledge data to obtain feature knowledge data matched with the retrieval tag, and then the tag probability corresponding to the retrieval tag is compared with the recommendation probability of the feature knowledge data matched with the retrieval tag.
And 105, if the label probability of the search label is larger than the recommendation probability of the feature knowledge data, taking the feature knowledge data as the knowledge to be recommended.
Specifically, if the tag probability of the search tag is greater than the recommendation probability of the feature knowledge data, then the feature knowledge data is taken as the knowledge to be recommended, and if the recommendation probability corresponding to one piece of feature knowledge data is 0.5, and the tag probability corresponding to the search tag matched with the piece of feature knowledge data is 0.6, at this time, the tag probability of the search tag is greater than the recommendation probability of the feature knowledge data, then the piece of feature knowledge data can be recommended, that is, the piece of feature knowledge data is taken as the knowledge to be recommended; if the recommendation probability corresponding to one piece of feature knowledge data is 0.6, and the label probability corresponding to the search label matched with the feature knowledge data is 0.5, at the moment, the label probability of the search label is not larger than the recommendation probability of the feature knowledge data, the recommendation is not performed, namely the feature knowledge data is not taken as knowledge to be recommended.
And step 106, recommending knowledge to the user based on the knowledge to be recommended.
Specifically, the knowledge to be recommended is recommended to the user according to the knowledge to be recommended obtained by comparing the label probability of the search label with the recommendation probability of the matched characteristic knowledge data. According to the embodiment of the invention, the related state-run assets supervision data can be intelligently matched and recommended according to the requirements and the query conditions of the user, the time and the cost of the user in the information retrieval process are reduced, the working efficiency and the decision accuracy of the user are improved, and the intelligent recommendation technology can help the user to acquire the required information more quickly by providing the data and the information related to the requirements of the user, so that the working efficiency and the decision accuracy are improved.
In an embodiment of the present invention, recommending knowledge to a user based on the knowledge to be recommended includes:
Dividing the search label into a plurality of label sets according to the label probability of the search label and a preset segmentation threshold value;
combining the knowledge to be recommended corresponding to the retrieval tag in the same tag set to obtain a knowledge set to be recommended;
And recommending knowledge to the user based on the knowledge set to be recommended.
Specifically, after knowledge to be recommended is obtained, the knowledge to be recommended is divided into a plurality of segments according to the label probability of the search labels matched with the knowledge to be recommended and the method of median or quartile extraction, a preset segment threshold is formed, the search labels are divided into a plurality of label sets according to the preset segment threshold, and then the knowledge to be recommended corresponding to the search labels in the same label set is combined to obtain the knowledge set to be recommended.
In an embodiment of the present invention, recommending knowledge to a user based on the knowledge set to be recommended includes:
Sequencing the knowledge sets to be recommended according to the segmentation sequence corresponding to the tag set to obtain a sequencing result of the knowledge sets to be recommended;
And recommending knowledge to the user according to the sequencing result of the knowledge set to be recommended.
Specifically, after the knowledge set to be recommended is obtained, the knowledge set to be recommended is sequenced according to the segmentation sequence corresponding to the label set, a sequencing result of the knowledge set to be recommended is obtained, and knowledge required by the user is recommended to the user in a personalized and intelligent mode according to the sequencing result of the knowledge set to be recommended.
According to the query conditions of the user, the embodiment of the invention intelligently retrieves state-run assets the supervision data and provides the data and information related to the user requirements. According to the query conditions provided by the user, based on the information provided by the user, extracting keywords, performing intelligent association on the labels, performing label classification, generating search labels, establishing association between the labels, and generating search labels and corresponding label probabilities of the user. And matching the characteristic knowledge data output by the probability knowledge model according to the retrieval tag of the user and the corresponding tag probability to obtain knowledge to be recommended, and combining the knowledge to be recommended according to the tag probability of the retrieval tag and the preset segmentation threshold value. And carrying out intelligent matching and retrieval on state-run assets supervision data by utilizing rules and relevance in the knowledge to be recommended and the retrieval tag, and carrying out sequencing and output of the knowledge set to be recommended according to the retrieval tag, and presenting data and information related to user requirements to the user, thereby improving the information acquisition efficiency and pertinence of the user.
In the embodiment of the invention, the probability knowledge model is obtained through training in the following way:
Acquiring service data and a type label corresponding to the service data;
Acquiring business knowledge and business rules;
Extracting characteristic knowledge data according to the service knowledge, the service rule, the service data and the type label corresponding to the service data to obtain the characteristic knowledge data and the type label corresponding to the characteristic knowledge data;
Acquiring initial probability aiming at the characteristic knowledge data;
training a preset probability knowledge model based on the feature knowledge data, the type label corresponding to the feature knowledge data and the initial probability to obtain a probability knowledge model.
Specifically, service data and a type tag corresponding to the service data are acquired first, and the service data can be state-run assets data related to service operation of a supervision platform, such as enterprise operation profit, net asset yield, asset liability and the like.
The business knowledge and business rules are acquired, and can be related knowledge and rules in state-run assets supervision fields, for example, state-run assets supervision standards, enterprise operation evaluation standards and the like, and can be realized by cooperation with state-run assets supervision departments, expert consultation, literature research and the like. Knowledge about the aspects of enterprise financial data, business indexes, industry classifications, etc. can be obtained by in-depth knowledge of state-run assets regulatory regulations, guidelines, etc.
According to business knowledge and business rules, knowledge extraction is carried out, characteristic knowledge data are extracted and are expressed in a form suitable for machine learning and reasoning, for example, key indexes and ratios in financial data can be extracted, and a correlation network among enterprises is established; matching the characteristic knowledge data with the type labels corresponding to the service data to obtain the characteristic knowledge data and the type labels corresponding to the characteristic knowledge data; combining the characteristic knowledge related data information according to the standard rule and the experience splitting rule to obtain characteristic knowledge data, type labels corresponding to the characteristic knowledge data and corresponding recommendation probability; and carrying out model training on the combined feature knowledge related data information to obtain feature knowledge data, type labels corresponding to the feature knowledge data and corresponding recommendation probability, wherein the model training is a dynamic process and needs to be updated and iterated continuously. In the training process, the combined data can be evaluated in a cross-validation mode, an index evaluation mode and the like so as to ensure the accuracy and the reliability of the combined data. By updating the data periodically, retraining the combined data, and evaluating the performance of the model. Meanwhile, the model can be improved and optimized according to new data and domain knowledge.
The method comprises the steps of obtaining initial probability aiming at characteristic knowledge data, constructing a preset probability knowledge model according to the characteristic knowledge data, type labels corresponding to the characteristic knowledge data, corresponding recommendation probability and expert knowledge in state-run assets supervision fields, and initializing experience values. The pre-set probabilistic knowledge model contains basic rules and knowledge that can be used to initially classify and annotate state-run assets the regulatory data. These initial empirical values may be provided by a practitioner or may be extracted from existing data. And training a preset probability knowledge model by using a machine learning algorithm. By using the previously acquired service data and the type label corresponding to the service data, the machine learning algorithm can automatically adjust the parameters and weights of a preset probability knowledge model according to the characteristics and the labels of the data, and the accuracy and generalization capability of the model are improved. The process of training the experience value can be realized by using methods such as supervised learning, unsupervised learning or semi-supervised learning. Training a preset probability knowledge model, and calculating through the probability model to obtain the probability knowledge model. The probability knowledge model is an intelligent model which is optimized and adjusted, and probability matching of different types of knowledge is carried out by setting and training a knowledge probability threshold value, so that data can be more accurately understood and applied state-run assets. The probabilistic model contains knowledge and rules learned from training data and updates to empirical values.
In the embodiment of the present invention, the acquiring service data and a type tag corresponding to the service data includes:
acquiring service data;
And identifying the service data based on a preset labeling model, determining the type of the service data and generating a type label.
Specifically, the service data is obtained by interfacing state-run assets the supervision platform, large model data, external related data files, etc., and the service data may be state-run assets data related to the operation of the supervision platform, for example, enterprise operation profits, net asset yields, asset liabilities, etc., that is, state-run assets data related to data indexes of the actual production operations of the supervision enterprise, and after the service data is obtained, the service data is preprocessed to form the specification data. The method comprises the steps of obtaining a preset labeling model, inputting service data to be preprocessed to form standard data to the preset labeling model, and enabling the preset labeling model to intelligently label the preprocessed data. The preset labeling model can automatically identify and label corresponding labels according to the content and the context of the business data, output labeled business data and type labels corresponding to the business data, for example, labels such as business income, net profit and the like can be labeled for financial data; for the enterprise name, labels such as "national enterprise", "marketing company" and the like can be attached.
In the embodiment of the invention, the preset labeling model is obtained by training in the following manner:
acquiring historical service data, wherein the historical service data comprises historical service knowledge and historical service rules;
And training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model.
Specifically, historical service data is obtained, the historical service data comprises historical service knowledge and historical service rules, the historical service data is preprocessed, and the initial labeling model is trained by utilizing technologies such as machine learning, natural language processing and the like. Training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model, wherein the model can learn and understand state-run assets relevant knowledge and rules in the supervision field, and meanings and relevance of financial data and operation indexes. Through training, the model can automatically identify and label different types of data. And after the model labeling is completed, verifying and correcting the labeling result by combining intelligent checking and manual checking of the machine learning model, and if labeling errors or omission are found, correcting and labeling again by using the model. After the labeling is completed, all final labeling data and relevant information are formed into a labeling data set and a labeling set to be output.
According to the embodiment of the invention, through the labeling model, state-run assets supervision data can be automatically classified and labeled. Therefore, the workload of manual marking can be reduced, and the efficiency of data processing is improved. The labeling model can rapidly and accurately label data, and label data is provided for subsequent processing and analysis, so that a large amount of time and labor cost are saved.
Based on the labeled data, a probabilistic knowledge model in state-run assets supervision fields is generated by using an inference algorithm and a probabilistic model. These probabilistic knowledge models can extract key knowledge and rules from the data and transform them into a form that state-run assets regulatory businesses can understand and apply. By applying the probability knowledge model, the accuracy and the reliability of data analysis can be improved, and the influence of human factors on the data analysis result can be reduced.
Based on the generated probability knowledge model, the embodiment of the invention can intelligently recommend the related state-run assets supervision data according to the requirements and the query conditions of the user, and help the user to quickly acquire the required information. The supervisory data can be intelligently retrieved state-run assets according to the query conditions of the user, providing data and information related to the user's needs. The intelligent functions can improve the working efficiency of users and reduce the time and cost of information retrieval.
The embodiment of the invention can improve the utilization value of state-run assets supervision data. The intelligent labeling technology can label the data, so that the data is easier to understand and apply. The probabilistic knowledge model can extract key knowledge and rules from the data to help users better understand and analyze the data. The intelligent recommending and intelligent searching functions can provide personalized and accurate data recommending and searching, and specific requirements of users are met. The application of the techniques can improve the utilization efficiency and the value of the data and help users to better utilize state-run assets to monitor the data for decision making and analysis. According to the embodiment of the invention, through improving the data processing efficiency, improving the data analysis accuracy, providing intelligent recommending and searching functions, improving the data utilization value and other effects, state-run assets supervision data can be effectively processed and analyzed, and the data utilization efficiency and value are improved.
In an embodiment of the present invention, the generating, according to the search requirement, a search tag and a corresponding tag probability includes:
Acquiring a preset demand conversion model;
And inputting the search requirement to the requirement conversion model to obtain the search label and the corresponding label probability.
Specifically, a preset demand conversion model is obtained, the search demand of the user is input to the demand conversion model, and the demand conversion model can output the search label corresponding to the search demand and the label probability corresponding to the search demand. The user search requirement refers to the description or expression of the required information by the user, and can be expressed by means of keywords, query conditions, questions and the like. The retrieval requirements of the users are processed and identified through keyword extraction, problem analysis and other methods, the retrieval requirements of the users are classified by using a classification algorithm or a machine learning model according to information provided by the users, and corresponding classification labels are generated. Attributes are extracted as labels from them using attribute extraction techniques. The attribute may be a specific attribute mentioned by the user in the description of the problem, such as time, place, price, etc., and the relationship between the identified and parsed tags is established by the methods of tag classification, attribute extraction, etc., and the relationship between the tags may be established by using techniques such as knowledge graph or association rule mining.
Referring to fig. 2, a schematic diagram of a recommendation method in an embodiment of the present invention is shown, in an embodiment of the present invention, taking state-run assets regulatory domain as an example, the recommendation method is described in an exemplary manner, and may specifically include the following steps:
and (5) intelligent labeling. And acquiring service data through an API interface, a database, a large model, related files and the like, and labeling the service data to obtain the service data and a label corresponding to the service data. And forming the final all labeling data and related information into a labeling data set and a labeling set. State-run assets the regulatory data is automatically annotated and classified by predefined rules and models. The system analyzes and processes the input state-run assets supervision data and related information according to preset labels and classification rules, automatically marks the input state-run assets supervision data and related information as corresponding categories or attributes, and generates a label set and a data set of state-run assets supervision data. Therefore, the workload and the error rate of manual marking can be reduced, and the accuracy and the efficiency of data processing are improved.
And (6) generating a probability knowledge model. And extracting the characteristic labels by extracting key knowledge and rules from the data, and extracting related knowledge content to generate a knowledge sequence. And extracting information such as relevance, regularity and importance by analyzing and mining marked data, assembling and generating a probability knowledge model according to state-run assets supervision standard rules and experience splitting rules, namely an initial knowledge model X in the graph, wherein the output of the probability knowledge model is characteristic knowledge data, labels and corresponding recommendation probabilities, the data M and the data N in the graph belong to the characteristic knowledge data, and the experience value X in the graph refers to the recommendation probability.
And (5) intelligent recommendation. Based on the generated initial knowledge model, the initial knowledge model is generated by analyzing the requirements and the query conditions of the user, the knowledge model is given to a training experience value by model training, the trained probability knowledge model, namely the experience knowledge model Y in the graph, is generated by probability model calculation based on the experience knowledge model, and the knowledge model and the experience knowledge model are personalized and intelligently matched with the requirements of the user. And according to the matching result, intelligently recommending relevant supervision measures and suggestions. The intelligent recommendation can intelligently recommend state-run assets supervision data according to an empirical knowledge model. For example, when a user needs to query financial data of a certain enterprise, the intelligent recommendation can intelligently generate a financial knowledge model of the certain enterprise from a database according to a query tag model of the enterprise, generate a financial management knowledge model and a probability model of the enterprise according to historical training experience values, and conduct directional recommendation according to the tag model. The intelligent recommendation can improve the working efficiency of the user and reduce the time and cost of information retrieval.
And (5) intelligent retrieval. And splitting the label according to the query condition of the user, and modeling and assembling the label to form a label model, namely, retrieving the label. And combining the tag model and the knowledge model according to the weights and the segmentation threshold, namely according to the tag probability of the retrieval tag and the preset segmentation threshold, and sequencing and outputting the knowledge model according to the user tag model. The embodiment of the invention realizes the intelligent retrieval function. The intelligent retrieval can utilize a knowledge model to intelligently retrieve state-run assets supervision data according to the query conditions of the user. The user can input keywords, conditions or problems, intelligent search can be performed according to the knowledge model to perform intelligent matching and search, and data and information related to the user requirements are returned. The intelligent retrieval can improve the information acquisition efficiency of the user, and help the user to better understand and analyze state-run assets the supervision data.
The embodiment of the invention realizes the processing and analysis of state-run assets supervision data through the technologies of intelligent labeling, probability knowledge model generation, intelligent recommendation, intelligent retrieval and the like. The techniques can automatically classify and label the data, generate a probability knowledge model for state-run assets supervision business understanding and application, provide intelligent recommendation and retrieval functions, and improve the utilization value and efficiency of state-run assets supervision data. State-run assets supervision departments can rapidly acquire information related to national asset supervision by using the recommendation method, and conduct personalized recommendation according to personal preference and historical behaviors, so that the working efficiency and decision accuracy are improved.
It should be noted that, for simplicity of description, the method embodiments are shown as a series of acts, but it should be understood by those skilled in the art that the embodiments are not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred embodiments, and that the acts are not necessarily required by the embodiments of the invention.
Referring to fig. 3, a block diagram of an embodiment of a recommendation device of the present invention is shown, and may specifically include the following modules:
A requirement acquisition module 201, configured to acquire a search requirement of a user;
A tag generation module 202, configured to generate a search tag and a corresponding tag probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag;
The knowledge generation module 203 is configured to obtain a probability knowledge model, and input the search tag to the probability knowledge model to obtain feature knowledge data and a corresponding recommendation probability;
the probability comparison module 204 is configured to match the search tag with the feature knowledge data, and compare a tag probability of the search tag with a recommendation probability of the matched feature knowledge data;
A determining module 205, configured to take the feature knowledge data as knowledge to be recommended if the tag probability of the search tag is greater than the recommendation probability of the feature knowledge data;
And the recommending module 206 is configured to recommend knowledge to the user based on the knowledge to be recommended.
In an embodiment of the present invention, the recommendation module includes:
The sub-splitting module is used for dividing the search tag into a plurality of tag sets according to the tag probability of the search tag and a preset segmentation threshold value;
the combination sub-module is used for combining the knowledge to be recommended corresponding to the retrieval tag in the same tag set to obtain a knowledge set to be recommended;
And the recommending sub-module is used for recommending knowledge to the user based on the knowledge set to be recommended.
In an embodiment of the present invention, the recommendation sub-module includes:
The sorting unit is used for sorting the knowledge sets to be recommended according to the segmentation sequence corresponding to the tag set to obtain a sorting result of the knowledge sets to be recommended;
And the recommending unit is used for recommending knowledge to the user according to the sequencing result of the knowledge set to be recommended.
In an embodiment of the present invention, the knowledge generation module includes:
The first acquisition sub-module is used for acquiring service data and type labels corresponding to the service data;
the second acquisition sub-module is used for acquiring service knowledge and service rules;
The characteristic knowledge determining submodule is used for extracting characteristic knowledge data according to the service knowledge, the service rule, the service data and the type label corresponding to the service data to obtain the characteristic knowledge data and the type label corresponding to the characteristic knowledge data;
A third obtaining sub-module, configured to obtain an initial probability for the feature knowledge data;
And the training sub-module is used for training a preset probability knowledge model based on the characteristic knowledge data, the type label corresponding to the characteristic knowledge data and the initial probability to obtain a probability knowledge model.
In an embodiment of the present invention, the first obtaining sub-module includes:
The acquisition unit is used for acquiring service data;
And the generating unit is used for identifying the service data based on a preset labeling model, determining the type of the service data and generating a type label.
In an embodiment of the present invention, the generating unit includes:
the acquisition subunit is used for acquiring historical service data, wherein the historical service data comprises historical service knowledge and historical service rules;
And the training subunit is used for training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model.
In an embodiment of the present invention, the tag generation module includes:
the demand conversion model acquisition sub-module is used for acquiring a preset demand conversion model;
and the search label generation sub-module is used for inputting the search requirement to the requirement conversion model to obtain the search label and the corresponding label probability.
The recommendation method provided by the embodiment of the invention needs to acquire the retrieval requirement of the user; then according to the search requirement, generating a search label and a corresponding label probability; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag; acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability; then matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data; if the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended; and recommending knowledge to the user based on the knowledge to be recommended. By the method, the search labels are generated according to the search requirements of the users and are input into the probability knowledge model, so that the problem that the keyword search algorithm in the prior art can only match accurate keywords and cannot process semantic relativity of synonyms, paraphraseology and the like can be solved, and the search accuracy is improved.
For the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments for relevant points.
As shown in fig. 4, an embodiment of the present invention further provides a structural block diagram of an electronic device, including:
the computer program comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein the computer program realizes the processes of the above recommended method embodiments when being executed by the processor, and can achieve the same technical effects, and the repetition is avoided, and the description is omitted here.
Fig. 5 shows a block diagram of a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the processes of the above-mentioned embodiment of the recommendation method, and achieves the same technical effects, so that repetition is avoided and no further description is given here.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.
It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the invention may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or terminal device that comprises the element.
The foregoing has outlined some of the more detailed description of the preferred method, apparatus, device and storage medium of the present invention, wherein specific examples are provided herein to illustrate the principles and embodiments of the present invention and to help understand the method and core idea of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.
Claims (10)
1. A recommendation method, the method comprising:
Acquiring the retrieval requirement of a user;
Generating a search label and a corresponding label probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag;
Acquiring a probability knowledge model, and inputting the retrieval tag to the probability knowledge model to obtain feature knowledge data and corresponding recommendation probability;
Matching the retrieval tag with the feature knowledge data, and comparing the tag probability of the retrieval tag with the recommendation probability of the matched feature knowledge data;
If the label probability of the search label is larger than the recommendation probability of the feature knowledge data, the feature knowledge data is used as the knowledge to be recommended;
and recommending knowledge to the user based on the knowledge to be recommended.
2. The method of claim 1, wherein recommending knowledge to a user based on the knowledge to be recommended comprises:
Dividing the search label into a plurality of label sets according to the label probability of the search label and a preset segmentation threshold value;
combining the knowledge to be recommended corresponding to the retrieval tag in the same tag set to obtain a knowledge set to be recommended;
And recommending knowledge to the user based on the knowledge set to be recommended.
3. The method of claim 2, wherein recommending knowledge to a user based on the set of knowledge to be recommended comprises:
Sequencing the knowledge sets to be recommended according to the segmentation sequence corresponding to the tag set to obtain a sequencing result of the knowledge sets to be recommended;
And recommending knowledge to the user according to the sequencing result of the knowledge set to be recommended.
4. The method of claim 1, wherein the probabilistic knowledge model is trained by:
Acquiring service data and a type label corresponding to the service data;
Acquiring business knowledge and business rules;
Extracting characteristic knowledge data according to the service knowledge, the service rule, the service data and the type label corresponding to the service data to obtain the characteristic knowledge data and the type label corresponding to the characteristic knowledge data;
Acquiring initial probability aiming at the characteristic knowledge data;
training a preset probability knowledge model based on the feature knowledge data, the type label corresponding to the feature knowledge data and the initial probability to obtain a probability knowledge model.
5. The method of claim 4, wherein the obtaining the service data and the type tag corresponding to the service data comprises:
acquiring service data;
And identifying the service data based on a preset labeling model, determining the type of the service data and generating a type label.
6. The method of claim 5, wherein the pre-set labeling model is trained by:
acquiring historical service data, wherein the historical service data comprises historical service knowledge and historical service rules;
And training the initial labeling model according to the historical service data and the preset label to obtain a trained labeling model.
7. The method of claim 1, wherein generating a search tag and a corresponding tag probability according to the search requirement comprises:
Acquiring a preset demand conversion model;
And inputting the search requirement to the requirement conversion model to obtain the search label and the corresponding label probability.
8. A recommendation device, the device comprising:
the demand acquisition module is used for acquiring the retrieval demand of the user;
The label generating module is used for generating a search label and a corresponding label probability according to the search requirement; the retrieval tag is used for describing retrieval requirements; the tag probability represents the probability that the user desires to use this retrieved tag;
The knowledge generation module is used for acquiring a probability knowledge model, inputting the retrieval tag to the probability knowledge model and obtaining characteristic knowledge data and corresponding recommendation probability;
the probability comparison module is used for matching the retrieval tag with the characteristic knowledge data and comparing the tag probability of the retrieval tag with the recommendation probability of the matched characteristic knowledge data;
the determining module is used for taking the characteristic knowledge data as knowledge to be recommended if the label probability of the search label is larger than the recommendation probability of the characteristic knowledge data;
and the recommending module is used for recommending knowledge to the user based on the knowledge to be recommended.
9. An electronic device, comprising: a processor, a memory and a computer program stored on the memory and capable of running on the processor, which when executed by the processor implements the steps of the recommendation method according to any of claims 1-7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the recommendation method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410108066.6A CN118069822A (en) | 2024-01-25 | 2024-01-25 | Recommendation method, recommendation device, recommendation equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410108066.6A CN118069822A (en) | 2024-01-25 | 2024-01-25 | Recommendation method, recommendation device, recommendation equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118069822A true CN118069822A (en) | 2024-05-24 |
Family
ID=91104972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410108066.6A Pending CN118069822A (en) | 2024-01-25 | 2024-01-25 | Recommendation method, recommendation device, recommendation equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118069822A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119441568A (en) * | 2025-01-08 | 2025-02-14 | 北京卓越未来国际医药科技发展有限公司 | Enterprise information retrieval method, device and electronic device based on knowledge engine |
-
2024
- 2024-01-25 CN CN202410108066.6A patent/CN118069822A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119441568A (en) * | 2025-01-08 | 2025-02-14 | 北京卓越未来国际医药科技发展有限公司 | Enterprise information retrieval method, device and electronic device based on knowledge engine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11829855B2 (en) | Time-factored performance prediction | |
| US11663254B2 (en) | System and engine for seeded clustering of news events | |
| CN103154991B (en) | Credit risk is gathered | |
| CN111767716B (en) | Method and device for determining enterprise multi-level industry information and computer equipment | |
| US12106351B1 (en) | System and method for automatic product matching | |
| CN113051365A (en) | Industrial chain map construction method and related equipment | |
| CN110188291B (en) | Document processing based on proxy log | |
| CN118941365B (en) | Commodity pushing method and system based on user preference analysis | |
| CN118657592B (en) | A personalized recommendation method and system for corporate financial products based on preference analysis | |
| CN110310012B (en) | Data analysis method, device, equipment and computer readable storage medium | |
| US11921754B2 (en) | Systems and methods for categorization of ingested database entries to determine topic frequency | |
| CN118861381A (en) | Recruitment information data processing method and system | |
| CN119226536A (en) | Financial business processing method and system, storage medium, and computer program product | |
| CN119378906A (en) | A task matching method based on user information and task characteristics | |
| CN118069822A (en) | Recommendation method, recommendation device, recommendation equipment and storage medium | |
| Nithya et al. | AI-Driven Legal Automation to Enhance Legal Processes with Natural Language Processing | |
| CN111104422A (en) | Training method, device, equipment and storage medium of data recommendation model | |
| CN117993876B (en) | Resume evaluation system, method, device and medium | |
| CN117540004B (en) | Industrial domain intelligent question-answering method and system based on knowledge graph and user behavior | |
| CN119128248A (en) | A method and system for intelligent job recommendation based on artificial intelligence | |
| Pawade et al. | Survey on resume and job profile matching system | |
| CN115293867A (en) | Financial reimbursement user portrait optimization method, device, equipment and storage medium | |
| CN112905790A (en) | Method, device and system for extracting qualitative indexes of supervision events | |
| CN119441499B (en) | Construction method, device and equipment of financial event map | |
| CN118964745B (en) | Government affair big data recommendation method, system, storage medium and electronic equipment |
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 |