CN110489649A - The method and device of label association content - Google Patents
The method and device of label association content Download PDFInfo
- Publication number
- CN110489649A CN110489649A CN201910764554.1A CN201910764554A CN110489649A CN 110489649 A CN110489649 A CN 110489649A CN 201910764554 A CN201910764554 A CN 201910764554A CN 110489649 A CN110489649 A CN 110489649A
- Authority
- CN
- China
- Prior art keywords
- content
- label
- word
- tag
- data
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/954—Navigation, e.g. using categorised browsing
-
- 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)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the present disclosure is related to Internet technical field, provides a kind of method and device of label association content, wherein method comprises determining that label to be associated；The semantic matches of the word label marked according to label to be associated and data content in data content as a result, determine that circle selects content；Content is selected to be associated with to label to be associated on circle.The embodiment of the present disclosure improves the efficiency and accuracy of label association content.
Description
Technical Field
The present disclosure relates to the field of internet technologies, and in particular, to a method and an apparatus for associating content with a tag.
Background
With the rapid development of internet technology, more and more contents can be obtained from the internet. And setting an operation label for guiding the user in the process of operating the website. After the operation label is determined, the stored data content is manually selected at the website server side according to the operation label, and the operation label is associated with the data content stored at the server side, so that the efficiency is low.
The data content types of the same website can be various, the data contents of different types belong to different services, and the form of the data contents can be greatly different. Different workers have different subjective awareness, are difficult to achieve uniformity in content selection and description of different types of data content, and are not enough to label one data content with sufficient operation labels.
Disclosure of Invention
In order to solve the above problems in the prior art, the present disclosure provides a scheme for tagging associated content.
According to an aspect of the embodiments of the present disclosure, a method for tagging associated content is provided, which includes: a label determining step, namely determining a label to be associated; a content circling step, namely determining circled content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content; and a content association step, namely associating the circled content to the label to be associated.
In one example, prior to the content circling step, the method further comprises: a content analyzing step, namely analyzing the data content to obtain a keyword matched with the data content; a label matching step, namely selecting word labels matched with the keywords from all the word labels; and a content labeling step, wherein the selected word label is used for labeling the data content.
In one example, the content circling step includes: a label analysis step, namely performing semantic analysis on the labels to be associated to obtain semantic participles; a label recommendation step, namely determining word labels with the similarity to the semantic word segmentation larger than or equal to a preset threshold; a label screening step, namely performing processing matched with an operation instruction on the word label based on the label to be associated according to the received operation instruction, wherein the operation instruction comprises a selection instruction and/or a logic operation instruction; and a content determining step, namely taking the data content marked by the word label obtained after the processing as the circled content.
In one example, the method further comprises: and a label changing step, namely changing the existing word label based on the keyword matched with the data content, wherein the change comprises a deleting operation and an adding operation.
In one example, the data content includes a teletext content, and the content parsing step includes: the image-text content disassembling step, namely disassembling the image-text content to obtain text content and image content; a text keyword obtaining step, namely performing semantic analysis and/or position importance analysis on text content to determine text keywords; and an image feature extraction step, wherein feature extraction is carried out on the image content based on the corpus data in the image corpus database to obtain image feature keywords.
In one example, the tag matching step includes: a label selection step, namely respectively selecting word labels matched with the text keywords and the image characteristic keywords; and a label combining step, wherein the word label matched with the text keyword and the same word label in the word labels matched with the image characteristic keywords are used as the word label for labeling the image-text content.
In one example, the method further comprises: and a content pushing step, namely pushing the content to be selected matched with the label to be associated to the user based on the frequency of using the label to be associated by the user within a preset time range.
In one example, the method further comprises: and a content releasing step, releasing the circled content to a content position matched with the label to be associated.
According to another aspect of the embodiments of the present disclosure, there is provided an apparatus for tagging associated content, including: the tag determining unit is used for determining tags to be associated; the content circle selection unit is used for determining circle selection content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content; and the content association unit is used for associating the circled content with the tag to be associated.
In one example, the apparatus further comprises: the content analysis unit is used for analyzing the data content to obtain a keyword matched with the data content; the label matching unit is used for selecting word labels matched with the keywords from all the word labels; and the content labeling unit is used for labeling the data content by using the selected word label.
In one example, the content circle selection unit includes: the tag analysis module is used for performing semantic analysis on the tags to be associated to obtain semantic participles; the tag recommendation module is used for determining word tags with the similarity degree with the semantic word segmentation larger than or equal to a preset threshold value; the tag screening module is used for performing processing matched with the operation instruction on the word tag based on the tag to be associated according to the received operation instruction, wherein the operation instruction comprises a selection instruction and/or a logic operation instruction; and the content determining module is used for taking the data content marked by the word label obtained after the processing as the circled content.
In one example, the apparatus further comprises: and the label changing unit is used for changing the existing word label based on the keyword matched with the data content, wherein the change comprises deletion operation and addition operation.
In one example, the data content includes a teletext content, and the content parsing unit includes: the image-text content disassembling module is used for disassembling the image-text content to obtain text content and image content; the text keyword acquisition module is used for performing semantic analysis and/or position importance analysis on text content to determine text keywords; and the image feature extraction module is used for extracting features of the image content based on the corpus data in the image corpus database to obtain image feature keywords.
In one example, the tag matching unit includes: the label selection module is used for respectively selecting word labels matched with the text keywords and the image characteristic keywords; and the label combining module is used for taking the word label matched with the text keyword and the same word label in the word labels matched with the image characteristic keywords as the word label for labeling the image-text content.
In one example, the apparatus further comprises: and the content pushing unit is used for pushing the content to be selected matched with the label to be associated to the user based on the frequency of using the label to be associated by the user within the preset time range.
In one example, the apparatus further comprises a content delivery unit for delivering the circled content to a content location matching the tag to be associated.
According to another aspect of the embodiments of the present disclosure, there is provided an electronic device, including:
a memory for storing a computer program;
a processor for executing the computer program stored in the memory, and when the computer program is executed, the method for associating content with tag of any of the above embodiments is implemented.
According to another aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium storing a computer program, where the computer program is executed by a processor to implement the method for tag-associated content of any of the above embodiments.
Based on the method and device for associating the content with the label, the electronic device and the computer readable storage medium, the unified word label is adopted to label various types of data content, and the data content is associated with the operation label after the word label is matched with the operation label, so that the efficiency and the accuracy of associating the data content with the operation label are improved.
Drawings
The above and other objects, features and advantages of the embodiments of the present disclosure will become readily apparent from the following detailed description read in conjunction with the accompanying drawings. Several embodiments of the present disclosure are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which:
FIG. 1 illustrates a flow diagram of one embodiment of a method of tagging associated content according to the present disclosure;
FIG. 2 illustrates a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure;
FIG. 3 illustrates a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure;
FIG. 4 illustrates a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure;
FIG. 5 illustrates a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure;
FIG. 6 illustrates a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure;
FIG. 7 illustrates a schematic structural diagram of one embodiment of an apparatus for tagging content in accordance with the present disclosure;
FIG. 8 shows a schematic structural diagram of another embodiment of an apparatus for tagging associated content according to the present disclosure;
FIG. 9 is a schematic diagram illustrating an embodiment of a content selection unit of an apparatus for tagging associated content according to the present disclosure;
FIG. 10 shows a schematic block diagram of another embodiment of an apparatus for tagging content in accordance with the present disclosure;
FIG. 11 is a schematic diagram illustrating an embodiment of a content parsing unit of the apparatus for tagging associated content according to the present disclosure;
FIG. 12 is a schematic diagram illustrating an embodiment of a tag matching unit of the tag-related-content apparatus according to the present disclosure;
FIG. 13 shows a schematic block diagram of another embodiment of an apparatus for tagging content in accordance with the present disclosure;
fig. 14 shows a schematic structural diagram of an embodiment of the electronic device of the present disclosure.
In the drawings, the same or corresponding reference numerals indicate the same or corresponding parts.
Detailed Description
The principles and spirit of the present disclosure will be described with reference to a number of exemplary embodiments. It is understood that these embodiments are given solely for the purpose of enabling those skilled in the art to better understand and to practice the present disclosure, and are not intended to limit the scope of the present disclosure in any way.
It should be noted that, although the expressions "first", "second", etc. are used herein to describe different modules, steps, data, etc. of the embodiments of the present disclosure, the expressions "first", "second", etc. are merely used to distinguish between different modules, steps, data, etc. and do not indicate a particular order or degree of importance. Indeed, the terms "first," "second," and the like are fully interchangeable.
The disclosed embodiments may be applied to electronic devices such as terminal devices, computer systems, and servers, which may operate with numerous other general purpose or special purpose computing system environments or configurations. Examples of well known terminal devices, computing systems, environments, and/or configurations that may be suitable for use with electronic devices, such as terminal devices, computer systems, and servers, include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, hand-held or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, networked personal computers, minicomputer systems, mainframe computer systems, distributed cloud computing environments that include any of the above, and the like.
Electronic devices such as terminal devices, computer systems, and servers may be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, and data structures, etc. that perform particular tasks or implement particular abstract data types. The computer system/server may be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media including memory storage devices.
In the mobile internet era, many companies use content as a traffic portal while achieving resource change using transactional means. Therefore, a content-plus-transaction platform generally includes various UGC, OGC, PGC contents, and various goods, services, and transactions. In some travel websites, the content can be divided into various forms such as notes, travel notes, strategies, questions and answers, and the like, and the commodity service also comprises various types such as free products, hotel air tickets and the like. Under such a large platform, how to make good interconversion between different types of contents, contents and commodities, becomes an important problem.
Because the types of the contents are numerous and belong to different service lines, the content types have great difference, one service line only operates for one type of contents, and in the description of the contents, the different service lines do not have a unified method. The efficiency of manual labeling of contents is low, and when the operators are increased, the subjective operation labels can be enlarged continuously, the manual labeling of all the contents is difficult, and meanwhile, the labeling of enough and sufficient labels on one content is difficult. In order to solve the above problem, embodiments of the present disclosure provide a scheme for associating content with a tag.
A first aspect of embodiments of the present disclosure provides a method for tagging associated content. FIG. 1 is a flow chart of one embodiment of a method for tagging content according to the present disclosure. As shown in fig. 1, the method of this embodiment includes: step 100-step 300. The method for associating the tag with the content in the present embodiment is described in detail below with reference to fig. 1.
Step 100, determining a label to be associated. In this embodiment, the tag to be associated may be a tag that attracts a user to set manually, may be a title of a note, a travel note, a strategy, and a question and answer, and may also be a tag that names a product, a brief description of a product action, and the like, which is not limited in this embodiment.
Step 200, determining circled content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content. The word label can be a label stored in a label database and used for labeling data content, and the form of the word label can be a single word or a word symbol with certain meaning such as a word or a phrase.
In this embodiment, the word labels stored in the label database may be stored after being divided step by step according to categories. For example, in the tag database of a travel website, the first level of word tags may include "travel time", "travel crowd", "travel mode", "travel preparation", "cate", "lodging", "traffic", "travel scene", "shopping", "entertainment item", and the like; second level word labels, such as "meal time", "meal service", "restaurant", etc., may also be included under the first level word label of "food; third level word labels, such as "michelin", "cafeteria", "themed restaurant", "roadside stand", "tea restaurant", etc., may also be included under the second level word label of "restaurant".
The word labels at all levels are related words around tourism, the word labels from coarse granularity to fine granularity can be classified from large to small, to subjects, to entity words and the like, and the word labels in the word label library can be used for depicting the content in detail granularity, so that different types of data contents such as notes, shorthand, strategies, questions and answers and the like are converted into an expression form consisting of multi-level labels. The word labels are managed according to the classification level, and corresponding word labels can be found more conveniently when the word labels in the label library are operated. And in the process that the labels to be associated are matched with the word labels, all the word labels in the label library are subjected to undifferentiated matching association, and the word labels with the similarity to be associated within a preset threshold range are searched in all the word labels. The operation that the labels to be associated are classified according to the classification level of the word labels is avoided, and the efficiency of associating the word labels to the labels to be associated is improved.
By using the same word label in the mode, different types of data contents can be labeled, and the defect that the data contents on different service lines are difficult to achieve unification and correlation can be effectively avoided. For example, the word label is "Chengdu", and can be used to label notes, travel notes, strategies, questions and answers of the aspects of Chengdu cate, Chengdu playing and the like, and can also be used to label "free-run" products on a travel website with the destination of Chengdu, and airplane tickets, train tickets, and the like with the destination of Chengdu.
The same data content may be tagged with one or more word tags. The purpose of labeling the data content by using the set unified data label is to unify the data content of various different types by using the data label.
The tags to be associated associate the data contents through the word tags, so that the sources of the data contents associated with the same tag to be associated can come from different service lines, the diversity of the types of the data contents associated with the tag to be associated is improved, and the utilization rate of the data contents is improved. The user can acquire more types of data contents through the same tag to be associated, and the use experience level of the user is improved.
And step 300, associating the circled content to the label to be associated.
The to-be-associated tags are matched with word tags in the tag library after semantic analysis, and word tags used for representing the semantics of the to-be-associated tags can be determined. The word label can determine the labeled data content. And associating the tag to be associated with the data content by taking the established word tag as a medium, so as to establish the association relationship between the tag to be associated and the data content.
In the method for associating tag with content provided by this embodiment, the tag library is established to store the word tags for tagging data content, and data content of each type can be tagged by the word tags in the tag library. So that the same word label can label different types of data content. The tag to be associated associates the corresponding data content through the word tag, so that the diversity of the data content associated with the tag to be associated is improved. The labor cost is reduced, and the efficiency and the accuracy of matching the data content with the tag to be associated are improved.
In one example, the data content may be labeled with hierarchical word labels, such as a tour. Zhangzhou ancient city distributes concentrated ancient early flavor, and with Fujian Zhangzhou, the ancient early flavor can be counterfeited in Taiwan. The ancient house with red bricks has a large number of storefronts with a horse-riding type, a Chinese-western combined type building, memorial archways left in the Ming and Qing era and strong ancient early tastes. The qi of people in ancient cities is strong, and there are dozens of ancient early-flavor snacks, Zhangzhou marinated noodles, triangular rice cakes, four-fruit soup, Zhangzhou fruit juiced juice and the like in the southern Fujian region, and many snacks in Taiwan actually come from the southern Fujian region. Zhangzhou ancient city, really can satisfy your all demands with one-stop, and the cheap price of things lets you 100 yuan just can eat up the wall and go out, both is historical ancient city, also eats goods auditorium! "in such a context, the label system of this example is used to de facto it. Firstly, segmenting data content, performing semantic analysis, position importance analysis and the like on the segmented data content to extract keywords of the data content, matching and screening the keywords and all word labels, and determining the word labels for labeling the data content. Because the word labels are managed according to the classification level, the word labels determined after the matching of the word labels to be associated is completed can be the word labels including the upper level of the word labels. For example, the word label in the aforementioned travel notes is finally determined as "food", "town", "Fujian", "Zhangzhou", "four fruit soup", "Zhangzhou braised noodles", "triangular rice cake" or the like by matching the keyword with the word label, and when the final word label is determined, the last word label of the aforementioned word label may be added, and the word label finally extracted may include "activities in the middle of the line: a delicious food; and (3) landscape: ancient town; destination: building in a Fujian province; destination: zhangzhou; topic: tp _ 548; topic: tp _ 153; POI: zhangzhou ancient city; the entity word: soup of four fruits; the entity word: zhangzhou marinated noodles; the entity word: triangular rice cake, etc. The problem that descriptions of different types of data contents are difficult to label by adopting a uniform label is solved.
In some embodiments, fig. 2 illustrates a flow diagram of another embodiment of a method of tagging associated content of the present disclosure. As shown in fig. 2, prior to step 200, the method further comprises steps 400-600. Wherein,
and step 400, analyzing the data content to obtain a keyword matched with the data content.
After the data content is submitted to the server, the data content can be analyzed in real time, so that word labels matched with the data content are found in the label database, and the data content is labeled. The data content can be conveniently found in time through the associated tag to be associated or the operation tag, and the utilization rate of the data content is improved.
In one example, the result of parsing the data content may be a keyword that can summarize the data content or the product.
And 500, selecting a word label matched with the keyword from all the word labels.
After the data content keywords are obtained, similarity matching is carried out on the keywords and the words or phrases of the word labels, and the words or phrases which are the same as the keywords are selected from all the word labels. And matching the keywords with the word labels by setting a similarity threshold. For example, word labels with similarity to the keywords of 0.8 may be used for labeling, and word labels smaller than a preset threshold may not be used for labeling the data content. The similarity between the keyword and the word label may be determined based on the number of the same words in the keyword and the word label. For example, the keyword is "Everest", and a similar word label is found to be "Everest". The number of the same words of the keyword and the word label is same as 2, the number of different words is diff equal to 3, and the length of the "Everest peak" is lena equal to 5. The keyword and word label similarity calculation formula may be: same/lens ≈ 0.4. The embodiment does not limit how to determine the similarity between the keywords and the word labels. The word labels used for marking the data content are matched through the data content keywords, and the efficiency of matching the word labels with the data content can be improved.
And step 600, labeling the data content by using the selected word label.
The word label selected to label the data content is obtained based on the similarity of the keywords of the data content and the word label. Tags that label data content may adequately summarize data content. Moreover, the data contents of the same type and the data contents of different types can be labeled by the same word label. The diversity of the content types of the word label labeling data can be improved.
In some embodiments, fig. 3 illustrates a flow diagram of another embodiment of a method of tagging associated content in the present disclosure. As shown in fig. 3, step 200, comprises step 201-step 204, wherein,
step 201, performing semantic analysis on the tags to be associated to obtain semantic participles.
The tags to be associated are artificially added for the operator to promote the corresponding data content, have certain subjectivity, and the subjective factors become more and more obvious as the operator continuously increases. In order to match accurate data content for the tag to be associated and avoid being influenced by subjective factors, the embodiment performs semantic analysis on the tag to be associated and determines semantic segmentation. The meaning expressed by the to-be-associated labels can be unified after semantic analysis, for example, the to-be-associated labels are respectively 'Shanghai mark' and 'Oriental pearl', and after the semantic analysis, unified semantic segmentation with the destination of Shanghai and the sight spot of Oriental pearl tower can be obtained.
In one example, the trained neural network can be used to perform semantic analysis on the labels to be associated. The method comprises the steps of using a label to be associated of an associated word label for a period of time as a training sample, and performing machine learning on an operation label, so that the operation label can be analyzed when the label to be associated is input, semantic word segmentation is obtained, and word labels are recommended automatically according to the semantic word segmentation, and therefore work of operators is greatly simplified when the label to be associated is associated.
Step 202, determining word labels with the similarity to the semantic word segmentation larger than or equal to a preset threshold. And analyzing the semantic participles obtained after the labels to be associated, matching the semantic participles with word labels in similarity, and selecting the word labels which are the same as the semantic participles or the word labels with the similarity larger than or equal to a preset threshold value.
Semantic participles are obtained after semantic analysis is performed on the tags to be associated, word tags expressing the tags to be associated are determined through the semantic participles, and accuracy of matching of the tags to be associated and the word tags is improved.
Step 203, according to the received operation instruction, processing matching with the operation instruction is carried out on the word label based on the label to be associated, wherein the operation instruction comprises a selection instruction and/or a logic operation instruction.
The word labels related in the embodiment are matched with the labels to be associated, and the obtained word labels are provided for the manufacturers of the labels to be associated to select, so that the work of operators in label association is greatly simplified. Wherein, the selected word label can be a word label which is finally reserved for use in the recommended word label; or may be a partial word tag that needs to be deleted in the recommended word tags. The logical operation instruction may include an intersect fetch, a union fetch, and a difference fetch. Taking two word labels A, B as an example, taking the intersection may be that the same data content circled according to the word labels is labeled by both a and B; the union set can be marked by A or B or marked by A and B in the same data content selected according to the word label; the difference set can be labeled by A but not labeled by B or labeled by B but not labeled by A in the same data content circled according to the word label.
Taking the above-mentioned pen as an example, such as "fujia snack gathering place", the operator associates data content for this operation label, and specifies a rule corresponding to the label as, a word label "destination: fujian, activities in the middle of the line: gourmet food ", thus keep the incidence relation of operation label and word label down. The step solves the problem of association between word labels and operation labels, so that technical word labels can be associated with the operated services through the system to generate specific service applications.
After the word labels are matched with the labels to be associated, operators can operate the matched word labels for formulating the rules, the work of meeting the same or similar data contents to be associated with the labels to be associated later is reduced, and the efficiency of the data contents to be associated with the labels to be associated is improved.
And step 204, taking the data content marked by the word label obtained after the processing as the circled content.
For the data content which is analyzed in real time, all types of data content are unified to the same dimension, including word labels of various levels and labels to be associated. Through the mode, the operator can select the data content which is desired and is matched with the word label according to the dimension of the operation topic.
The embodiment can select data contents more specifically by selecting and performing logical operation on the word labels matched with the labels to be associated.
Besides selecting word labels matched with the labels to be associated and selecting data contents after logical operation, the method also supports selection according to conditions such as release time, data content length, picture quantity and the like, and also supports various screening conditions for combining certain word labels, so that the required targeted data contents can be screened out.
Fig. 4 shows a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure. As shown in fig. 4, the method for associating tag with content of the present disclosure further includes step 700, modifying the existing word tag based on the keyword matched with the data content. Wherein, the change comprises a deletion operation and an addition operation.
The word labels stored in the label library for marking the data content can be modified according to the writing mode, writing words and other factors of the data content issued by the user. For example, with the development of network technology, people often use network terms in the writing process, for example, shanghai may be madam, beijing may be huddle, etc., and word tags in the tag library may be changed with keywords matching data contents. And changing, wherein the changing comprises deleting operation and adding operation. The method can also be a modification operation, specifically, deleting the word tags in the tag library and then adding new word tags.
The word labels in the label library can be changed according to a preset period, for example, the word labels in the label library are changed every month; for example, 48000 of 50000 randomly extracted travel notes with "blocked city" representing beijing may be changed according to the frequency of use of the keyword obtained by analyzing the data content, the word label "beijing" in the label library is deleted, and then the word label "blocked city" is added, or the word label "blocked city" is directly added on the premise of keeping "beijing".
By changing the word label, the accurate marking of the word label can be carried out on the data contents of various writing habits and writing styles.
Fig. 5 shows a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure. As shown in fig. 5, when the data content includes teletext content, step 400 in the method for tagging associated content according to the present disclosure may include steps 401 to 403, and step 500 may include steps 501 and 502, where,
step 401, disassembling the image-text content to obtain a text content and an image content. Different data content forms need different ways to extract keywords so as to determine word labels used for labeling the image-text contents.
Step 402, performing semantic analysis and/or position importance analysis on the text content to determine text keywords.
And training the model to be trained by adopting the travel notes to obtain a semantic model. Each segmented sentence of the note is used as an input, the important unimportant is used as a classification label, a word sequence (a text or a sentence) is input, and the probability that the word sequence belongs to different classification labels is output. And segmenting words and phrases in the sentence to form a feature vector, mapping the feature vector to the middle layer through linear transformation, and mapping the middle layer to the label. The non-linear activation function is used in predicting the label and is not used in the middle layer. For example, 10000 notes are selected as training samples, and whether each segmentation sentence in the samples is important is manually marked. And training the model to be trained. Learning and training are carried out based on the context and semantic content, and a semantic importance model training is carried out to obtain a semantic model which judges whether the segmentation sentence is important according to the input segmentation sentence.
Predicting the important probability and unimportant probability of the segmentation sentences by using the semantic model obtained after training, inputting each segmentation sentence to be analyzed and shorthand into the semantic model for prediction, and obtaining the important probability P of each segmentation sentencefast_posAnd a probability of unimportance Pfast_negBased on the importance probability P of the segmented sentencefast_posAnd a probability of unimportance Pfast_negCalculating a semantic importance score of the sentence, which may be obtained by dividing, subtracting, or other operation methods of the above two, and is not particularly limited herein, for example, the score may be Pfast_pos/Pfast_negMay also be Pfast_pos+Pfast_negAnd the like.
Since the shorthand expresses the complete meaning through the sentence, the extraction range of the keyword can be narrowed in the process of extracting the keyword by determining the semantic importance of the segmented sentence, namely, the keyword can be extracted from the segmented sentence with higher semantic importance score.
Will be in dataThe segmented sentences of the container, the semantic importance scores and the position marks of the segmented sentences, namely the chapter numbers, the paragraph numbers and the sequence numbers of the segmented sentences, are used as input features and trained to obtain a position importance model. The model is used for predicting the importance of the statement in the travel notes, and an important probability P can be obtainedxgb_posAnd a probability P of unimportancexgb_negUsing Pxgb_pos/Pxgb_neg(or other calculation methods, not limited herein) as the importance score of the final segmented sentence.
According to the embodiment, the summarization of important information of the article is improved through the text keywords obtained after semantic analysis and/or position importance analysis are carried out on the text content, and the subject content of the article is more accurately expressed.
Step 403, performing feature extraction on the image content based on the corpus data in the image corpus database to obtain image feature keywords.
In order to summarize the image content by using characters and determine the keywords matched with the image content, the pixel characteristics of the picture can be used in the process of extracting the image characteristics, and the results of objects or scenes in the picture, such as restaurants, lakes, plates, puppies and the like, are obtained through an ImageNet pre-training model. The results of a plurality of pictures of one content are respectively obtained to obtain texts matched with the image content, and keywords are extracted from the texts matched with the image content in the step 402.
Step 501, selecting word labels matched with the text keywords and the image characteristic keywords respectively. The keywords matching with the content of the image and text are obtained through the steps 402 and 403. The embodiment determines that the word labels are matched with the text content part and the image content part respectively based on the obtained keywords.
Step 502, using the word label matched with the text keyword and the same word label in the word labels matched with the image characteristic keywords as the word label for labeling the image-text content. In one example, word tags matching text content may be stored in a set a, word tags matching image content may be stored in a set b, both sets a and b may use word tags as elements, and word tags corresponding to image-text content may be word tags in an intersection of the sets a and b.
Fig. 6 shows a flow diagram of another embodiment of a method of tagging associated content according to the present disclosure. As shown in fig. 6, the method for associating content with a tag of the present disclosure may further include steps 800 and 900, where step 800 is to push content to be circled matching the tag to be associated to the user based on the frequency of using the tag to be associated by the user within a preset time range.
In one example, a user often browses a strategy that a tag to be associated is 'what you play' in Beijing within a certain time period, and after a certain frequency is reached, the server can push other types of data content associated with the tag to be associated to the user, so that the preference of the user is met, and the use experience of the user on a travel platform is improved.
And 900, delivering the circled content to a content position matched with the label to be associated. The content delivery task issues tasks, records an adjusted data content list, and sorts the data contents in real time according to different service requirements, including time sorting, content quality sorting, praise collection reply number sorting and the like. For the same task, different service lines can output data content according to different sequences, and the output mode comprises two modes of technical interface docking and operating Excel export. For the technical interface mode, automatic content delivery can be completed, and delivery to a specific content position can be directly output.
The embodiment solves the problems that various types of data contents associated with the tags to be associated cannot be unified and the operation labeling is imperfect and low in efficiency, and greatly improves the efficiency of content selection and delivery.
The embodiment of the disclosure can also count the content of various content types in the data content associated with the same tag to be associated. For example, how many notes, how many travel notes, how many questions and answers, etc., and how many contents each word label can cover can be analyzed, and similarly, the labels to be associated, the release time, the length of the labels to be associated, etc., can be subjected to uniform perspective analysis, so that the overall situation of the circled contents can be visually known, and whether the contents meet the requirements of the position to be put can be evaluated.
If the circled content is determined to be basically consistent with the label to be associated, the establishment of the task can be confirmed. And if the temporary task does not meet the requirement, deleting the temporary task and performing circle selection again. For the determined task, the selection condition of the task is stored in the background, the service line for creating the task and the creator information are recorded, and a task number is generated. The results of the tasks are updated according to the update frequency set by the creator, and one-time tasks are selected, or the tasks are updated by hours, days and the like.
For the established task, the operation can intervene in the task content in the content auditing platform, including operations of deleting a piece of content from the task, adjusting word labels or operation labels of the content, adjusting titles and content of the content, and the like, and for the updated content, the updated content is marked as unverified content, so that the following links cannot be entered. The results of the manually intervened task are saved and stored, and in later use, if the same task needs to be used, the task results can be directly used without re-intervention.
Based on the same inventive concept, a second aspect of the present disclosure provides an apparatus for tagging associated content, and uses the apparatus to implement the steps in the method for tagging associated content according to the first aspect and the embodiments.
FIG. 7 illustrates a schematic structural diagram of one embodiment of an apparatus for tagging content in accordance with the present disclosure; as shown in fig. 7, the apparatus for associating content with a tag of the present disclosure includes: a tag determination unit 10, configured to determine a tag to be associated; a content circle selection unit 20, configured to determine circle selection content in the data content according to a semantic matching result between the tag to be associated and the word tag labeled by the data content; a content associating unit 30, configured to associate the circled content with the tag to be associated.
In this embodiment, the tag to be associated may be a tag that attracts a user to set manually, may be a title of a note, a travel note, a strategy, and a question and answer, and may also be a tag that names a product, a brief description of a product action, and the like, which is not limited in this embodiment. The word label can be a label stored in a label database and used for labeling data content, and the form of the word label can be a single word or a word symbol with certain meaning such as a word or a phrase. The same data content may be tagged with one or more word tags. The purpose of labeling the data content by using the set unified data label is to unify the data content of various different types by using the data label.
The content association unit 30 associates the data content with the tag to be associated through the word tag, so that the data content associated with the same tag to be associated comes from different service lines, the diversity of the types of the data content associated with the tag to be associated is improved, and the utilization rate of the data content is improved. The user can acquire more types of data contents through the same tag to be associated, and the use experience level of the user is improved.
By establishing a tag library to store word tags for tagging data contents, data contents of various types can be tagged through the word tags in the tag library. So that the same word label can label different types of data content. The tag to be associated associates the corresponding data content through the word tag, so that the diversity of the data content associated with the tag to be associated is improved. The efficiency and the accuracy of the data content matched by the tag to be associated are improved.
Fig. 8 is a schematic structural diagram of another embodiment of the apparatus for tag-associated content according to the present disclosure, and as shown in fig. 8, the apparatus for tag-associated content according to the present embodiment further includes: a content analysis unit 40, configured to analyze the data content to obtain a keyword matched with the data content; a tag matching unit 50 for selecting a word tag matching the keyword among all the word tags; and a content labeling unit 60 for labeling the data content with the selected word label.
The content parsing unit 40 may parse the data content uploaded to the server in real time, so as to find the word tag matching the data content in the tag database, and label the data content. The result obtained by analyzing the data content by the content analyzing unit 40 may be a keyword capable of summarizing the data content or a product, so that the data content can be conveniently found in time through the associated tag to be associated or the operation tag, and the utilization rate of the data content is improved.
After acquiring the data content keyword, the tag matching unit 50 selects a word or a phrase identical to the keyword from all the word tags by similarity matching between the keyword and the word or the phrase of the word tag. And matching the keywords with the word labels by setting a similarity threshold. For example, word labels with similarity to the keywords of 0.8 may be used for labeling, and word labels smaller than a preset threshold may not be used for labeling the data content. The similarity between the keyword and the word label may be determined based on the number of the same words in the keyword and the word label.
The word label selected by the content labeling unit 60 to label the data content is obtained based on the similarity of the keyword of the data content and the word label. Tags that label data content may adequately summarize data content. Moreover, the data contents of the same type and the data contents of different types can be labeled by the same word label. The diversity of the content types of the word label labeling data can be improved.
Fig. 9 is a schematic structural diagram illustrating an embodiment of a content circling unit of the apparatus for tagging associated content according to the present disclosure. As shown in fig. 9, the content circle selection unit 20 includes: the tag analysis module 21 is configured to perform semantic analysis on the to-be-associated tags to obtain semantic segmentation; the label recommending module 22 is used for determining word labels with the similarity degree with the semantic word segmentation larger than or equal to a preset threshold value; the tag screening module 23 is configured to perform processing on the word tag matched with the operation instruction based on the tag to be associated according to the received operation instruction, where the operation instruction includes a selection instruction and/or a logic operation instruction; and the content determining module 24 is configured to use the data content labeled by the processed word label as the circled content.
In order to match accurate data content for the tag to be associated and avoid being influenced by subjective factors, the tag parsing module 21 of this embodiment performs semantic analysis on the tag to be associated and determines semantic segmentation. The meanings expressed by the labels to be associated can be unified through semantic analysis.
The tag recommendation module 22 analyzes the semantic participle obtained after the tag to be associated is analyzed, performs similarity matching between the semantic participle and the word tag, and selects the word tag which is the same as the semantic participle or the word tag with the similarity greater than or equal to a preset threshold. Semantic participles are obtained after semantic analysis is performed on the tags to be associated, word tags expressing the tags to be associated are determined through the semantic participles, and accuracy of matching of the tags to be associated and the word tags is improved.
The word label selected by the label screening module 23 may be a word label finally reserved for use among the recommended word labels; or may be a partial word tag that needs to be deleted in the recommended word tags. The logical operation instruction may include an intersect fetch, a union fetch, and a difference fetch. Taking two word labels A, B as an example, taking the intersection may be that the same data content circled according to the word labels is labeled by both a and B; the union set can be marked by A or B or marked by A and B in the same data content selected according to the word label; the difference set can be labeled by A but not labeled by B or labeled by B but not labeled by A in the same data content circled according to the word label.
Besides selecting word labels matched with the labels to be associated and selecting data contents after logical operation, the method also supports selection according to conditions such as release time, data content length, picture quantity and the like, and also supports various screening conditions for combining certain word labels, so that the required targeted data contents can be screened out.
For data content that has been parsed in real time, the content determination module 24 unifies all types of data content into the same dimension, including word tags of various levels and tags to be associated. Through the mode, the operator can select the data content which is desired and is matched with the word label according to the dimension of the operation topic.
FIG. 10 shows a schematic block diagram of another embodiment of an apparatus for tagging content in accordance with the present disclosure; as shown in fig. 10, the apparatus for associating content with a tag of this embodiment further includes: and a tag changing unit 70, configured to change an existing word tag based on the keyword matched with the data content, where the change includes a deletion operation and an addition operation.
In this embodiment, the word tag for labeling the data content stored in the tag library may be modified according to the writing manner, the writing words, and other factors of the data content issued by the user. The word labels in the label library can be changed according to a preset period, and the change comprises deletion operation and addition operation. The method can also be a modification operation, specifically, deleting the word tags in the tag library and then adding new word tags. For example, the word tags in the tag library are changed every month; or the use frequency of the keywords obtained after the data content analysis can be changed. By changing the word label, the accurate marking of the word label can be carried out on the data contents of various writing habits and writing styles.
In some embodiments, the data content includes teletext content, and fig. 11 shows a schematic structural diagram of an embodiment of a content parsing unit of the apparatus for tagging associated content according to the present disclosure. As shown in fig. 11, the content analysis unit 40 includes: the image-text content disassembling module 41 is configured to disassemble the image-text content to obtain text content and image content; a text keyword obtaining module 42, configured to perform semantic analysis and/or position importance analysis on text content, and determine a text keyword; and an image feature extraction module 43, configured to perform feature extraction on image content based on the corpus data in the image corpus database to obtain an image feature keyword.
In order to ensure the accuracy and the integrity of the extraction of the keywords of the data content, different data content forms need to be subjected to keyword extraction in different ways so as to determine word labels for labeling the image-text content. And training the model to be trained by adopting the travel notes to obtain the trained semantic model. The trained semantic model is used for performing semantic analysis and/or position importance analysis on text content to determine text keywords. Each segmented sentence of the note is used as an input, the important unimportant is used as a classification label, a word sequence (a text or a sentence) is input, and the probability that the word sequence belongs to different classification labels is output. And segmenting words and phrases in the sentence to form a feature vector, mapping the feature vector to the middle layer through linear transformation, and mapping the middle layer to the label. The non-linear activation function is used in predicting the label and is not used in the middle layer. According to the embodiment, the summarization of important information of the article is improved through the text keywords obtained after semantic analysis and/or position importance analysis are carried out on the text content, and the subject content of the article is more accurately expressed.
In order to summarize the image content by using characters and determine the keywords matched with the image content, the pixel characteristics of the picture can be used in the process of extracting the image characteristics, and the results of objects or scenes in the picture, such as restaurants, lakes, plates, puppies and the like, are obtained through an ImageNet pre-training model. The results of the several pictures of one content are respectively obtained to obtain texts matched with the image content, and the texts matched with the image content are used for extracting keywords through a text keyword obtaining module 42.
FIG. 12 is a schematic diagram illustrating an embodiment of a tag matching unit of the tag-related-content apparatus according to the present disclosure; as shown in fig. 12, the tag matching unit of the present embodiment includes: a tag selection module 51, configured to select word tags matching the text keywords and the image feature keywords respectively; and the label combining module 52 is configured to use the word label matched with the text keyword and the same word label in the word labels matched with the image feature keywords as the word label for labeling the image-text content.
The data content is determined to be a text keyword by the text keyword obtaining module 42, the image feature extracting module 43 is used for obtaining an image feature keyword, and the label selecting module 51 is used for determining that the text content part and the image content part are matched with the word label respectively based on the obtained keyword. In one example, word tags matching text content may be stored in a set a, word tags matching image content may be stored in a set b, both sets a and b may use word tags as elements, word tags corresponding to image content may be stored in a set a in an intersection of the sets a and b, and word tag combination module 52 may store word tags matching text content in a set a and store word tags matching image content in a set b, both sets a and b may use word tags as elements, and word tags corresponding to image content may be word tags in an intersection of the sets a and b.
FIG. 13 shows a schematic block diagram of another embodiment of an apparatus for tagging content in accordance with the present disclosure; as shown in fig. 13, the apparatus for tag-associated content of this embodiment further includes a content pushing unit 80, configured to push, to the user, content to be circled that matches the tag to be associated, based on the frequency of using the tag to be associated by the user within the preset time range. The user often browses the same or similar tags to be associated within a certain time period, and after a certain frequency is reached, the content pushing unit 80 can push other types of data content associated with the tags to be associated to the user, so as to cater to the user preference, and improve the use experience of the user on the travel platform.
With continued reference to fig. 13, as shown in fig. 13, the apparatus for associating a tag with content of the present embodiment further includes a content delivery unit 90, configured to deliver the circled content to a content location matching the tag to be associated. And the content delivery task issues tasks, records the adjusted data content list, and sorts the data contents in real time according to different service requirements, including time sorting, content quality sorting, compliment collection reply number sorting and the like. For the same task, different service lines can output data content according to different sequences, and the output mode comprises two modes of technical interface docking and operating Excel export.
By adopting the device for associating the label with the content in any embodiment to operate the label with the content, the problems that various types of data content to be associated with the label cannot run through the same label and the operation labeling is imperfect and low in efficiency are solved, and the efficiency of content selection and delivery is greatly improved.
Fig. 14 shows a schematic structural diagram of an embodiment of the electronic device of the present disclosure. Referring now to fig. 14, shown is a schematic diagram of an electronic device suitable for use in implementing a terminal device or server of an embodiment of the present application. As shown in fig. 14, the electronic device includes a processor and a memory. The electronic device may also include input and output means. The memory and the input/output device are connected with the processor through the bus. The memory is used for storing instructions executed by the processor; and the processor is used for calling the instructions stored in the memory and executing the method for associating the label with the content related to the embodiment.
In the embodiment of the disclosure, the processor can call an instruction stored in the memory to determine the tag to be associated; determining circled content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content; and associating the circled content with the label to be associated. The process of executing the tag associated content by the electronic device may refer to the implementation process of the method for associating the tag associated content described in the foregoing embodiment, and is not described herein again.
The embodiment of the disclosure also provides a computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are run on a computer, the method for associating content with a tag according to the embodiment is executed.
The embodiment of the present disclosure further provides a computer program product containing instructions, which when run on a computer, causes the computer to execute the method for associating content with a tag according to the embodiment.
In one or more optional implementation manners, the embodiment of the disclosure further provides a computer-readable storage medium for storing computer-readable instructions, and the instructions when executed enable a computer to execute the method for associating the content by the tag in any one of the possible implementation manners. In another alternative example, the computer program product is embodied as a Software product, such as a Software Development Kit (SDK) or the like.
Although operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in serial order, or that all illustrated operations be performed, to achieve desirable results. In certain environments, multitasking and parallel processing may be advantageous.
The methods and apparatus of the present disclosure can be accomplished with standard programming techniques with rule-based logic or other logic to accomplish the various method steps. It should also be noted that the words "means" and "module," as used herein and in the claims, is intended to encompass implementations using one or more lines of software code, and/or hardware implementations, and/or equipment for receiving inputs.
Any of the steps, operations, or procedures described herein may be performed or implemented using one or more hardware or software modules, alone or in combination with other devices. In one embodiment, the software modules are implemented using a computer program product comprising a computer readable medium containing computer program code, which is executable by a computer processor for performing any or all of the described steps, operations, or procedures.
The foregoing description of the implementations of the disclosure has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the disclosure. The embodiments were chosen and described in order to explain the principles of the disclosure and its practical application to enable one skilled in the art to utilize the disclosure in various embodiments and with various modifications as are suited to the particular use contemplated.
Claims (11)
1. A method for tagging associated content, comprising:
a label determining step, namely determining a label to be associated;
a content circling step, namely determining circled content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content;
and a content association step, namely associating the circled content to the label to be associated.
2. The method of claim 1, wherein prior to the content circling step, the method further comprises:
a content analyzing step, analyzing the data content to obtain a keyword matched with the data content;
a label matching step, namely selecting a word label matched with the keyword from all word labels;
and a content labeling step, in which the selected word label is used for labeling the data content.
3. The method of claim 2, wherein the content circling step comprises:
a label analyzing step, namely performing semantic analysis on the label to be associated to obtain semantic participles;
a label recommendation step, namely determining word labels with the similarity degree with the semantic word segmentation larger than or equal to a preset threshold value;
a label screening step, namely performing processing matched with the operation instruction on the word label based on the label to be associated according to the received operation instruction, wherein the operation instruction comprises a selection instruction and/or a logic operation instruction;
and a content determining step, namely taking the data content marked by the word label obtained after the processing as the circled content.
4. The method of claim 2 or 3, wherein the method further comprises:
and a label changing step, namely changing the existing word label based on the keyword matched with the data content, wherein the change comprises deleting operation and adding operation.
5. The method of claim 2, wherein the data content comprises teletext content, and the content parsing step comprises:
the image-text content disassembling step, namely disassembling the image-text content to obtain text content and image content;
a text keyword obtaining step, namely performing semantic analysis and/or position importance analysis on the text content to determine a text keyword;
and an image feature extraction step, wherein feature extraction is carried out on the image content based on the corpus data in the image corpus database to obtain image feature keywords.
6. The method of claim 5, wherein the tag matching step comprises:
a label selection step, namely respectively selecting word labels matched with the text keywords and the image characteristic keywords;
and a label combining step, wherein the word label matched with the text keyword and the same word label in the word labels matched with the image characteristic keywords are used as the word label for marking the image-text content.
7. The method of claim 1, wherein the method further comprises:
and a content pushing step, namely pushing the circled content matched with the to-be-associated label to the user based on the frequency of using the to-be-associated label by the user within a preset time range.
8. The method of claim 1, wherein the method further comprises:
and a content delivery step, delivering the circled content to a content position matched with the label to be associated.
9. An apparatus for tagging associated content, comprising:
the tag determining unit is used for determining tags to be associated;
the content circle selection unit is used for determining circle selection content in the data content according to the semantic matching result of the tag to be associated and the word tag marked by the data content;
and the content association unit is used for associating the circled content with the tag to be associated.
10. An electronic device, comprising:
a memory for storing a computer program;
a processor for executing a computer program stored in the memory, and when executed, implementing the method of tag-associated content of any of claims 1-8.
11. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method of tagging associated content according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910764554.1A CN110489649B (en) | 2019-08-19 | 2019-08-19 | Method and device for associating content with tag |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910764554.1A CN110489649B (en) | 2019-08-19 | 2019-08-19 | Method and device for associating content with tag |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110489649A true CN110489649A (en) | 2019-11-22 |
| CN110489649B CN110489649B (en) | 2023-06-27 |
Family
ID=68551500
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910764554.1A Active CN110489649B (en) | 2019-08-19 | 2019-08-19 | Method and device for associating content with tag |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110489649B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111462281A (en) * | 2020-03-31 | 2020-07-28 | 北京创鑫旅程网络技术有限公司 | Poster generation method, device, equipment and storage medium |
| CN112699237A (en) * | 2020-12-24 | 2021-04-23 | 百度在线网络技术（北京）有限公司 | Label determination method, device and storage medium |
| CN114298236A (en) * | 2021-12-30 | 2022-04-08 | 南京中孚信息技术有限公司 | Method, device and electronic device for determining similarity of unstructured content |
| CN114492419A (en) * | 2022-04-01 | 2022-05-13 | 杭州费尔斯通科技有限公司 | Text labeling method, system and device based on newly added key words in labeling |
| CN114817542A (en) * | 2022-04-29 | 2022-07-29 | 北京创鑫旅程网络技术有限公司 | Information identification method, apparatus, device, medium, and program product |
| CN115659969A (en) * | 2022-12-13 | 2023-01-31 | 成方金融科技有限公司 | Document labeling method and device, electronic equipment and storage medium |
| CN118551769A (en) * | 2024-07-25 | 2024-08-27 | 浙江鸟潮供应链管理有限公司 | Label circling method and device based on large language model and computer equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102750280A (en) * | 2011-04-19 | 2012-10-24 | 国际商业机器公司 | Computer processing method and system for search |
| CN102982076A (en) * | 2012-10-30 | 2013-03-20 | 新华通讯社 | Multi-dimensionality content labeling method based on semanteme label database |
| CN105354265A (en) * | 2015-10-23 | 2016-02-24 | 北京京东尚科信息技术有限公司 | Method and apparatus for automatically constructing association structure of delivered keyword |
| CN105843877A (en) * | 2016-03-21 | 2016-08-10 | 郑州悉知信息科技股份有限公司 | Page search method and apparatus |
-
2019
- 2019-08-19 CN CN201910764554.1A patent/CN110489649B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102750280A (en) * | 2011-04-19 | 2012-10-24 | 国际商业机器公司 | Computer processing method and system for search |
| CN102982076A (en) * | 2012-10-30 | 2013-03-20 | 新华通讯社 | Multi-dimensionality content labeling method based on semanteme label database |
| CN105354265A (en) * | 2015-10-23 | 2016-02-24 | 北京京东尚科信息技术有限公司 | Method and apparatus for automatically constructing association structure of delivered keyword |
| CN105843877A (en) * | 2016-03-21 | 2016-08-10 | 郑州悉知信息科技股份有限公司 | Page search method and apparatus |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111462281A (en) * | 2020-03-31 | 2020-07-28 | 北京创鑫旅程网络技术有限公司 | Poster generation method, device, equipment and storage medium |
| CN111462281B (en) * | 2020-03-31 | 2023-06-13 | 北京创鑫旅程网络技术有限公司 | Poster generation method, device, equipment and storage medium |
| CN112699237A (en) * | 2020-12-24 | 2021-04-23 | 百度在线网络技术（北京）有限公司 | Label determination method, device and storage medium |
| CN114298236A (en) * | 2021-12-30 | 2022-04-08 | 南京中孚信息技术有限公司 | Method, device and electronic device for determining similarity of unstructured content |
| CN114492419A (en) * | 2022-04-01 | 2022-05-13 | 杭州费尔斯通科技有限公司 | Text labeling method, system and device based on newly added key words in labeling |
| CN114492419B (en) * | 2022-04-01 | 2022-08-23 | 杭州费尔斯通科技有限公司 | Text labeling method, system and device based on newly added key words in labeling |
| CN114817542A (en) * | 2022-04-29 | 2022-07-29 | 北京创鑫旅程网络技术有限公司 | Information identification method, apparatus, device, medium, and program product |
| CN115659969A (en) * | 2022-12-13 | 2023-01-31 | 成方金融科技有限公司 | Document labeling method and device, electronic equipment and storage medium |
| CN118551769A (en) * | 2024-07-25 | 2024-08-27 | 浙江鸟潮供应链管理有限公司 | Label circling method and device based on large language model and computer equipment |
| CN118551769B (en) * | 2024-07-25 | 2024-11-12 | 浙江鸟潮供应链管理有限公司 | Tag selection method, device and computer equipment based on large language model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110489649B (en) | 2023-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110489649A (en) | The method and device of label association content | |
| CN110134931B (en) | Medium title generation method, medium title generation device, electronic equipment and readable medium | |
| CN107766371B (en) | Text information classification method and device | |
| CN104834729B (en) | Topic recommends method and topic recommendation apparatus | |
| CN112733042B (en) | Recommendation information generation method, related device and computer program product | |
| CN111104526A (en) | Financial label extraction method and system based on keyword semantics | |
| CN107832414A (en) | Method and apparatus for pushed information | |
| KR20170093713A (en) | Method and device for mobile searching based on artificial intelligence | |
| CN114065750B (en) | Product information matching and publishing method and its device, equipment, medium, and product | |
| CN107832338B (en) | Method and system for recognizing core product words | |
| CN107305551A (en) | The method and apparatus of pushed information | |
| CN103559220B (en) | Picture searching equipment, method and system | |
| CN111444304A (en) | Search ranking method and device | |
| CN109582788A (en) | Comment spam training, recognition methods, device, equipment and readable storage medium storing program for executing | |
| CN112749330A (en) | Information pushing method and device, computer equipment and storage medium | |
| US20240086452A1 (en) | Tracking concepts within content in content management systems and adaptive learning systems | |
| CN110287341A (en) | A kind of data processing method, device and readable storage medium storing program for executing | |
| CN112784156A (en) | Search feedback method, system, device and storage medium based on intention recognition | |
| CN107632974A (en) | Suitable for multi-field Chinese analysis platform | |
| CN112464106B (en) | Object recommendation method and device | |
| CN113204697A (en) | Searching method, searching device, electronic equipment and storage medium | |
| Kaur et al. | A Review on Sentimental Analysis on Facebook Comments by using Data Mining Technique | |
| Joseph et al. | Exploring the Application of Natural Language Processing for Social Media Sentiment Analysis | |
| US20240370498A1 (en) | Automatic textual document evaluation using large language models | |
| CN115168637B (en) | Method, system and storage medium for adding label to picture |
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 |