CN116244496B - Resource recommendation method based on industrial chain - Google Patents
Resource recommendation method based on industrial chain Download PDFInfo
- Publication number
- CN116244496B CN116244496B CN202211555850.9A CN202211555850A CN116244496B CN 116244496 B CN116244496 B CN 116244496B CN 202211555850 A CN202211555850 A CN 202211555850A CN 116244496 B CN116244496 B CN 116244496B
- Authority
- CN
- China
- Prior art keywords
- keyword
- keywords
- recommended
- resource information
- resource
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000010801 machine learning Methods 0.000 claims description 15
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 abstract description 5
- 238000007726 management method Methods 0.000 description 19
- 230000009286 beneficial effect Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration 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
- G06F16/90—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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a resource recommendation method based on an industrial chain. Comprising the following steps: acquiring operation information when a user performs a resource request on a current industrial chain, and extracting keywords to form a first initial keyword label; acquiring resource information to be recommended of a resource management background, extracting keywords of each resource information to be recommended, and forming a second keyword tag database; matching similar keywords with keywords in the first initial keyword label, and combining the keywords with the first initial keyword label to obtain a first combined keyword label; based on the first combined keyword labels and each second keyword label, matching the first combined keyword labels and each second keyword label one by one, and calculating a matching index of the corresponding second keyword label; and extracting a third keyword label with the matching index larger than a preset threshold value, and recommending the corresponding resource information. Through extraction and comparison of keywords, resource recommendation based on an industrial chain is achieved, recommendation to users can be more accurate, and requirements of users on resources are more efficient.
Description
Technical Field
The invention relates to the technical field of resource recommendation, in particular to a resource recommendation method based on an industrial chain.
Background
At present, the information in the Internet age is continuously expanded, and in various portal websites and information release platforms, the update of the content becomes the most basic problem of the websites. With the limitations of existing capabilities, how to cope with more information content has become an important point.
However, the information volume is huge and the update speed is high, and under the condition of the existing capacity limitation, the requirement of users on timely and accurate information recommendation needs is difficult to meet. Therefore, the problems of untimely user information recommendation and inaccurate resource recommendation in the industry chain are the primary problems to be considered.
Therefore, the invention provides a resource recommendation method based on an industrial chain.
Disclosure of Invention
The invention provides a resource management method based on an industrial chain, which is used for realizing resource recommendation based on the industrial chain through extraction and comparison of keywords, so that the resource recommendation of a user can be more accurate, and meanwhile, the sequential recommendation is performed according to the relative degree of the resources, and the user can acquire the resources more efficiently.
The invention provides a resource management method based on an industrial chain, which comprises the following steps:
step 1: acquiring operation information when a user makes a resource request on a current industrial chain, taking the operation information as comparison resource information, and extracting keywords in the comparison resource information to form a first initial keyword label;
step 2: acquiring resource information to be recommended of a resource management background, extracting keywords of each resource information to be recommended, and forming a second keyword tag database;
step 3: identifying and matching a plurality of keywords in the first initial keyword labels to obtain corresponding first similar keyword labels, and combining the corresponding first similar keyword labels with the first initial keyword labels to obtain first combined keyword labels;
step 4: based on the first combined keyword label and each second keyword label in the second keyword label database, matching the first combined keyword label and each second keyword label one by one, and calculating a matching index of the corresponding second keyword label;
step 5: and extracting a third keyword label with the matching index larger than a preset threshold value, and recommending the resource information matched with the third keyword label.
In one possible implementation manner, the obtaining the operation information when the user makes the resource request on the current industry chain as the reference resource information, and extracting the keywords in the reference information to form a first initial keyword tag includes:
acquiring all operation information when a target user makes a resource request in a preset time range on a current industrial chain, and taking the operation information as control resource information;
extracting words from all analysis sentences in the comparison resource information to obtain a first word set of the comparison resource information;
acquiring word characteristics of each word of the first word set and sentence characteristics of each word in a corresponding analysis sentence;
and extracting keywords from the comparison resource information based on the word characteristics and the sentence characteristics on the basis of a machine learning model to obtain all keywords of the comparison resource information, and further obtaining a first initial keyword label.
In one possible implementation manner, the extracting keywords from the comparison resource information based on the machine learning model and the word features and the sentence features to obtain all keywords of the current comparison resource information, and further obtain a first initial keyword tag includes:
extracting historical keywords in all historical control resource information of the resource management background and the criticality of each historical keyword, and training an initial learning model to obtain a machine learning model;
analyzing word characteristics of each word of a first word set of current comparison resource information and sentence characteristics of corresponding words in corresponding analysis sentences based on the machine learning model, acquiring the criticality of each first word in the first word set, eliminating the same keywords to obtain residual keywords and obtaining first initial keyword labels;
the first initial keyword tag comprises an operation serial number of corresponding operation information and a word sequence number of the rest keywords related to the operation information.
In one possible implementation manner, the obtaining resource information to be recommended in the resource management background, extracting keywords of each resource information to be recommended, and forming a second keyword tag database includes:
acquiring all effective resource information of the resource management background to obtain a plurality of resource information to be recommended;
calculating the relevance of each word in each analysis statement of each resource information to be recommended and the rest words in the analysis statement, so as to obtain the weight coefficient of each word in the current analysis statement, and creating a weight graph of each resource information to be recommended according to the weight coefficients of all words in the current analysis statement and the weight coefficient of the current analysis statement;
in the process of recommending historical resource information, the position of the corresponding vertex in the weighted graph is used as the probability of recommending keywords, and is used as the vertex recommending probability of each vertex in the corresponding weighted graph;
sequencing all vertex recommendation probabilities in each weighted graph, extracting second keywords in the corresponding resource information to be recommended, and obtaining corresponding second keyword labels;
a second keyword tag database is constructed based on all of the second keyword tags.
In one possible implementation manner, the identifying and matching of similar keywords are performed on a plurality of keywords in the first initial keyword label to obtain a corresponding first similar keyword label, and the corresponding first similar keyword label is combined with the first initial keyword label to obtain a first combined keyword label, which includes:
matching similar keywords corresponding to each first keyword in the first initial keyword label from a keyword matching database, so as to obtain a first similar keyword label;
and carrying out label combination on the first initial keyword label and the first similar keyword label to obtain a first combined keyword label.
In one possible implementation manner, the calculating the matching index of the corresponding second keyword tag based on the matching of the first merged keyword tag with each second keyword tag in the second keyword tag database one by one includes:
matching each second keyword label in the database of the first combined keyword labels and each second keyword label one by one, and determining a keyword matching index of each matching result;
；
wherein,keyword matching indexes corresponding to the matching results;refers to the inclusion of the first merged keyword tagThe first merged keywordWords and phrases;refers to the inclusion of the corresponding second keyword tag h in the second keyword tag databaseThe first keywordWords and phrases;、the part-of-speech conversion coefficient of the current word;、meaning the semantic conversion coefficient of the current word; a1 represents part-of-speech weights; a2 represents semantic weights.
In one possible implementation manner, the extracting a third keyword tag with a matching index greater than a preset threshold, and recommending the resource information matched with the third keyword tag includes:
obtaining a matching index of the first combined keyword tag and the current keyword tag of the second keyword tag database, and comparing the matching index with a preset threshold;
if the matching index is larger than a preset threshold value, extracting a second keyword label corresponding to the current matching index in the second keyword label database as a third keyword label;
performing resource recommendation based on the resource information matched with the current third keyword label;
and if the matching index is at the boundary value of the preset matching range, judging whether the current information to be recommended is recommended or not based on the information number of the information to be recommended.
In one possible implementation manner, if the matching index is at a boundary value of a preset matching range, determining whether to recommend the current information to be recommended based on the information number of the information to be recommended includes:
determining the number of the resource information to be recommended, which is obtained by acquiring the boundary between the matching indexes of all the resource information to be recommended and the contrast resource information in the resource management background and is in a preset matching range;
simultaneously, the number of the keywords of the first combined keyword tag corresponding to the comparison resource information is obtained;
if the number of the resource information to be recommended is larger than the preset multiple of the number of the keywords in the first combined keyword tag, judging that the current resource information to be recommended is not recommended;
if the number of the resource information to be recommended is not more than the preset multiple of the number of the keywords of the first combined keyword tag, judging that the current information to be recommended is recommended.
In one possible implementation manner, after extracting the third keyword label with the matching index greater than the preset threshold, and recommending the resource information matched with the third keyword label, the method further includes:
acquiring all third keyword labels, carrying out the sequence from high to low on the matching index of each third keyword label to obtain the arrangement sequence of the corresponding third keyword labels, and carrying out corresponding sequence recommendation on the information to be recommended based on the current sequence;
the information to be recommended corresponding to the third keyword label with the highest matching index is located at the first position of the current industrial chain recommendation position.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
The technical scheme of the invention is further described in detail through the drawings and the embodiments.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a flowchart of a resource management method based on an industrial chain according to an embodiment of the present invention;
FIG. 2 is a flowchart of a resource management method based on an industrial chain to obtain a first initial keyword tag according to an embodiment of the present invention;
fig. 3 is a flowchart of a method for constructing a second keyword tag database according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.
Example 1:
the embodiment of the invention provides a resource management method based on an industrial chain, as shown in fig. 1, comprising the following steps:
step 1: acquiring operation information when a user makes a resource request on a current industrial chain, taking the operation information as comparison resource information, and extracting keywords in the comparison resource information to form a first initial keyword label;
step 2: acquiring resource information to be recommended of a resource management background, extracting keywords of each resource information to be recommended, and forming a second keyword tag database;
step 3: identifying and matching a plurality of keywords in the first initial keyword labels to obtain corresponding first similar keyword labels, and combining the corresponding first similar keyword labels with the first initial keyword labels to obtain first combined keyword labels;
step 4: based on the first combined keyword label and each second keyword label in the second keyword label database, matching the first combined keyword label and each second keyword label one by one, and calculating a matching index of the corresponding second keyword label;
step 5: and extracting a third keyword label with the matching index larger than a preset threshold value, and recommending the resource information matched with the third keyword label.
In this embodiment, the operation information is actual operation information based on searching or clicking of resources by the target user on the current industry chain.
In this embodiment, the reference resource information is based on the sum of all the operation information of the target users in the current industry chain.
In this embodiment, the keywords against the resource information are commonly decided based on the word characteristics of each word and the sentence characteristics of each word in the corresponding analysis sentence.
In this embodiment, the first initial keyword tag includes an operation serial number of the corresponding operation information to which the first initial keyword tag belongs and a word sequence number of a keyword related to the operation information; for example, keywords include: searching, method and industry chain, wherein the corresponding first initial keyword label comprises the following steps: 10123 011 111 112, wherein 10123 is the operation serial number of the corresponding operation information, 011, 111, 112 is the word sequence number of the related keyword, wherein the first number represents the part of speech of the current keyword, such as 0 represents the verb, and 1 represents the noun.
In this embodiment, the resources to be recommended are based on all the resource information in the resource management background within the valid time.
In this embodiment, the second keyword tag database is constructed based on all of the second keyword tags.
In this embodiment, the second keyword tag is configured based on a keyword extracted from information of a resource to be recommended in the resource management background.
In this embodiment, the first similar keyword label is based on each keyword in the current first initial label, and the similar words corresponding to each keyword are screened in the database, and all similar words form a first similar keyword label; for example, the current first initial keyword tag includes: the method for obtaining the corresponding first similar keyword label comprises the following steps: primary school, middle school, university, mode, regimen.
In this embodiment, the first merged keyword tag is a keyword tag obtained by merging the first initial keyword tag and the first similar keyword tag.
In this embodiment, the matching index is a matching index for keyword matching based on each of the second keyword tags and the first merged keyword tag in the second keyword tag database.
In this embodiment, the third keyword tag means that if the matching index is greater than a preset threshold, the second keyword tag corresponding to the current matching index in the second keyword tag database is extracted as the third keyword tag.
The beneficial effects of the technical scheme are as follows: through extraction and comparison of keywords, resource recommendation based on an industrial chain is achieved, so that resource recommendation to a user can be more accurate, and meanwhile, sequential recommendation is conducted according to the degree of resource correlation, and the user can acquire the resources more efficiently.
Example 2:
based on embodiment 1, the obtaining the operation information of the current industrial chain when the user makes the resource request, as the comparison resource information, and extracting the keywords in the comparison resource information to form a first initial keyword tag includes:
acquiring all operation information when a target user makes a resource request in a preset time range on a current industrial chain, and taking the operation information as control resource information;
extracting words from all analysis sentences in the comparison resource information to obtain a first word set of the comparison resource information;
acquiring word characteristics of each word of the first word set and sentence characteristics of each word in a corresponding analysis sentence;
and extracting keywords from the comparison resource information based on the word characteristics and the sentence characteristics on the basis of a machine learning model to obtain all keywords of the comparison resource information, and further obtaining a first initial keyword label.
In this embodiment, the users on the current industry chain may be plural, but the target user refers to one user performing information operation on the current industry chain.
In this embodiment, the operation information is based on actual operation information of searching and clicking resources by the target user on the current industry chain and display information on a corresponding interface after the clicking performs the frame skip.
In this embodiment, the reference resource information is obtained by integrating all operation information of the target user in the preset time range on the current industrial chain.
In this embodiment, the analysis sentence is a sentence having a real meaning in the current collation resource information, and the analysis sentence is a sentence having no real meaning, such as an exclamation sentence composed of exclamation words, is not included.
In this embodiment, the first word set is a set of words extracted from all analysis sentences currently collating the resource information.
In this embodiment, for example, the term features include specific part-of-speech semantic features of the current term, and the sentence features include features such as sentence effects of the current term on the corresponding analysis sentence effect conditions.
In this embodiment, the machine learning model is obtained by training the initial learning model based on all histories in the resource management background against the history keywords in the resource information and the criticality of each history keyword.
In this embodiment, the extraction of the keywords is decided based on the result of comprehensively considering the word characteristics of the words in each analysis sentence and the sentence characteristics of the word influencing the sentence in comparison with each analysis sentence in the resource information.
In this embodiment, the first initial keyword tag includes an operation serial number of the corresponding operation information to which the first initial keyword tag belongs and a word sequence number of a keyword related to the operation information; for example, keywords include: searching, method and industry chain, wherein the corresponding first initial keyword label comprises the following steps: 10123 011 111 112, wherein 10123 is the operation serial number of the corresponding operation information, 011, 111, 112 is the word sequence number of the related keyword, wherein the first number represents the part of speech of the current keyword, such as 0 represents the verb, and 1 represents the noun.
The beneficial effects of the technical scheme are as follows: the method comprises the steps of extracting words in the operation information of the target user, determining keywords based on word characteristics and sentence characteristics of the words, obtaining keywords of the operation information corresponding to the current target user, and recommending resources according to the keywords, so that the resource recommendation is more accurate.
Example 3:
based on embodiment 2, the machine learning model is based on the word feature and the sentence feature, and the keyword extraction is performed on the comparison resource information to obtain all keywords of the comparison resource information, so as to obtain a first initial keyword tag, as shown in fig. 2, including:
extracting historical keywords in all historical control resource information of the resource management background and the criticality of each historical keyword, and training an initial learning model to obtain a machine learning model;
analyzing word characteristics of each word of a first word set of current comparison resource information and sentence characteristics of corresponding words in corresponding analysis sentences based on the machine learning model, acquiring the criticality of each first word in the first word set, eliminating the same keywords to obtain residual keywords and obtaining first initial keyword labels;
the first initial keyword tag comprises an operation serial number of corresponding operation information and a word sequence number of the rest keywords related to the operation information.
In this embodiment, the history-reference resource information is resource information corresponding to the operation performed by the history user within the effective time of the current resource management background.
In this embodiment, the history keyword is all keywords extracted from the history-reference resource information.
In this embodiment, the criticality of the history keywords is based on an average of the importance of each keyword in the analysis statement corresponding to the current collation resource information and the importance of the entire collation resource information at the current time.
In this embodiment, for example, the term features include specific part-of-speech semantic features of the current term, and the sentence features include features such as sentence effects of the current term on the corresponding analysis sentence effect conditions.
In this embodiment, the first initial keyword tag includes an operation serial number of the corresponding operation information to which the first initial keyword tag belongs and a word sequence number of the remaining keywords related to the operation information.
The beneficial effects of the technical scheme are as follows: the machine learning model is obtained by analyzing the historical keywords, and the keywords of the words obtained by the current comparison of the resource information are extracted based on the machine learning model, so that the keywords of the operation information corresponding to the current target user are obtained, and the resource recommendation can be more accurate according to the keyword resource recommendation.
Example 4:
based on the embodiment 1, the obtaining the resource information to be recommended in the resource management background, and extracting the keywords of each resource information to be recommended, to form a second keyword tag database, as shown in fig. 3, includes:
acquiring all effective resource information of the resource management background to obtain a plurality of resource information to be recommended;
calculating the relevance of each word in each analysis statement of each resource information to be recommended and the rest words in the analysis statement, so as to obtain the weight coefficient of each word in the current analysis statement, and creating a weight graph of each resource information to be recommended according to the weight coefficients of all words in the current analysis statement and the weight coefficient of the current analysis statement;
in the process of recommending historical resource information, the position of the corresponding vertex in the weighted graph is used as the probability of recommending keywords, and is used as the vertex recommending probability of each vertex in the corresponding weighted graph;
sequencing all vertex recommendation probabilities in each weighted graph, extracting second keywords in the corresponding resource information to be recommended, and obtaining corresponding second keyword labels;
a second keyword tag database is constructed based on all of the second keyword tags.
In this embodiment, the effective resource information is resource information that can be subjected to public query recommendation in an effective time.
In this embodiment, the relevance refers to the degree of relevance between the current word and other words.
In this embodiment, the weight coefficient of the word refers to the weight coefficient of the influence degree of the word on the analysis sentence in the analysis sentence corresponding to the current word.
In this embodiment, the weight coefficient of the analysis statement is a weight coefficient of the influence degree of the analysis statement on the resource to be recommended in the corresponding resource to be recommended information of the current analysis statement.
In this embodiment, the weight map of the resource to be recommended is composed of a weight coefficient of each word in the analysis sentence, a weight coefficient of each analysis sentence in the salary of the resource to be recommended, and each word.
In this embodiment, the recommendation probability of the vertex of the weighted graph is the probability of the current word serving as the keyword to recommend the resource.
In this embodiment, the second keyword in the corresponding resource information to be recommended is extracted by extracting, based on the ranking of all vertices of the weighted graph according to the recommendation probability, a word corresponding to a vertex with a recommendation probability greater than a preset value as the second keyword, thereby forming a second keyword label.
In this embodiment, the second keyword tag database is formed by the second keyword tags of all the resource information to be recommended.
The beneficial effects of the technical scheme are as follows: by judging the relevance of the words and the weight coefficient of the words in the corresponding analysis statement, the second keyword label corresponding to the current to-be-recommended resource information is obtained, and the second keyword label is constructed, so that the keywords of the current to-be-recommended resource can be obtained more accurately, the matching of the keywords is more accurate, and the resource recommendation is more accurate.
Example 5:
based on embodiment 3, the identifying and matching of similar keywords are performed on a plurality of keywords in the first initial keyword label to obtain a corresponding first similar keyword label, and the corresponding first similar keyword label is combined with the first initial keyword label to obtain a first combined keyword label, which includes:
matching similar keywords corresponding to each first keyword in the first initial keyword label from a keyword matching database, so as to obtain a first similar keyword label;
and carrying out label combination on the first initial keyword label and the first similar keyword label to obtain a first combined keyword label. For example, the first initial keyword tag has 3 keywords, the first similar keyword tag contains 5 keywords similar to the keywords in the first initial keyword tag, the first initial keyword tag and the first similar keyword tag are combined, namely 8 keywords contained in the two tags are combined to form a new set, then the similar keywords corresponding to the current keywords are arranged according to the sequence of the previous keywords, and after the arrangement of one group of keywords and the similar keywords is finished, the next keywords are arranged, so that the first combined keyword is obtained.
In this embodiment, the similar keywords are words that are identical/semantically similar to the current keyword part of speech contained in the keyword matching database.
The beneficial effects of the technical scheme are as follows: by performing similar matching on keywords in the first initial keyword tag extracted from the comparison resource information, keywords similar to the current keywords can be obtained, so that a first combined keyword tag is obtained, matching on the keywords can be more comprehensive, and more resource information to be recommended, which is matched with the comparison resource information, can be obtained.
Example 6:
based on embodiment 5, the matching index of the corresponding second keyword tag is calculated based on the matching of the first merged keyword tag with each second keyword tag in the second keyword tag database one by one, which includes:
matching each second keyword label in the database of the first combined keyword labels and each second keyword label one by one, and determining a keyword matching index of each matching result;
；
wherein,keyword matching indexes corresponding to the matching results;refers to the inclusion of the first merged keyword tagThe first merged keywordWords and phrases;refers to the inclusion of the corresponding second keyword tag h in the second keyword tag databaseThe first keywordWords and phrases;、meaning part of speech conversion of the current wordCoefficients;、meaning the semantic conversion coefficient of the current word; a1 represents part-of-speech weights; a2 represents semantic weights.
In this embodiment, the keyword matching index is a matching index for performing keyword matching based on each of the second keyword tags and the first merged keyword tag in the second keyword tag database.
In this embodiment, the keyword matching index is determined based on the part-of-speech matching degree and the semantic matching degree between keywords and the influence of the part-of-speech and the semantic on the words.
In this embodiment, the number of keywords of each second keyword tag may be different.
In the embodiment, the part-of-speech conversion coefficient and the semantic conversion coefficient are mainly used for converting corresponding words into the parameter values which can be calculated, so that keyword matching indexes can be conveniently calculated.
In this embodiment, the part-of-speech weight and the semantic weight are determined based on the part-of-speech and the degree of influence of the semantics on the term comparison process.
The beneficial effects of the technical scheme are as follows: and matching each second keyword label of the second keyword label database corresponding to the to-be-recommended resource information and the first combined keyword label corresponding to the to-be-recommended resource information one by one to obtain a matching index, so that the to-be-recommended resource information is recommended according to the matching index, and the resource recommendation can be more accurate.
Example 7:
based on embodiment 6, extracting a third keyword tag with a matching index greater than a preset threshold, and recommending resource information matched with the third keyword tag, including:
obtaining a matching index of the first combined keyword tag and the current keyword tag of the second keyword tag database, and comparing the matching index with a preset threshold;
if the matching index is larger than a preset threshold value, extracting a second keyword label corresponding to the current matching index in the second keyword label database as a third keyword label;
performing resource recommendation based on the resource information matched with the current third keyword label;
and if the matching index is at the boundary value of the preset matching range, judging whether the current information to be recommended is recommended or not based on the information number of the information to be recommended.
In this embodiment, the preset threshold is predetermined based on the historical recommendation information, and may be adjusted according to the actual recommendation condition.
In this embodiment, the third keyword tag means that if the matching index is greater than a preset threshold, the second keyword tag corresponding to the current matching index in the second keyword tag database is extracted as the third keyword tag.
The beneficial effects of the technical scheme are as follows: by recommending the resource information with the matching index larger than the preset threshold, the target user can obtain the recommended information with higher matching degree with the current operation information more accurately.
Example 8:
based on embodiment 7, if the matching index is at the boundary value of the preset matching range, determining whether the current information to be recommended is recommended based on the information number of the information to be recommended includes:
determining the number of the resource information to be recommended, which is obtained by acquiring the boundary between the matching indexes of all the resource information to be recommended and the contrast resource information in the resource management background and is in a preset matching range;
simultaneously, the number of the keywords of the first combined keyword tag corresponding to the comparison resource information is obtained;
if the number of the resource information to be recommended is larger than the preset multiple of the number of the keywords in the first combined keyword tag, judging that the current resource information to be recommended is not recommended;
if the number of the resource information to be recommended is not more than the preset multiple of the number of the keywords of the first combined keyword tag, judging that the current information to be recommended is recommended.
In this embodiment, the preset multiple may be adaptively adjusted according to the number of keywords.
The beneficial effects of the technical scheme are as follows: by carrying out classified discussion on the matching indexes at the boundary of the matching range, if the number of the resource information to be recommended is small, the resource information to be recommended in the boundary range is recommended, and if the number of the resource information to be recommended is large, the resource information to be recommended in the boundary range is abandoned, so that the recommended resource obtained by the target user is as accurate as possible and the requirement of the target user is met.
Example 9:
based on embodiment 7, after extracting the third keyword label with the matching index greater than the preset threshold, and recommending the resource information matched with the third keyword label, the method further includes:
acquiring all third keyword labels, carrying out the sequence from high to low on the matching index of each third keyword label to obtain the arrangement sequence of the corresponding third keyword labels, and carrying out corresponding sequence recommendation on the information to be recommended based on the current sequence;
the information to be recommended corresponding to the third keyword label with the highest matching index is located at the first position of the current industrial chain recommendation position.
In this embodiment, the third keyword tag means that if the matching index is greater than a preset threshold, the second keyword tag corresponding to the current matching index in the second keyword tag database is extracted as the third keyword tag.
In this embodiment, the recommendation order of the information to be recommended is arranged in order of the matching index of the corresponding third keyword tag from high to low.
The beneficial effects of the technical scheme are as follows: by comparing the matching indexes of the keyword labels and sorting the information to be recommended based on the comparison result of the matching indexes, the target user of the current industrial chain can acquire the resources more efficiently and rapidly.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (7)
1. The resource recommendation method based on the industrial chain is characterized by comprising the following steps of:
step 1: acquiring operation information when a user makes a resource request on a current industrial chain, taking the operation information as comparison resource information, and extracting keywords in the comparison resource information to form a first initial keyword label;
step 2: acquiring resource information to be recommended of a resource management background, extracting keywords of each resource information to be recommended, and forming a second keyword tag database;
step 3: identifying and matching a plurality of keywords in the first initial keyword labels to obtain corresponding first similar keyword labels, and combining the corresponding first similar keyword labels with the first initial keyword labels to obtain first combined keyword labels;
step 4: based on the first combined keyword label and each second keyword label in the second keyword label database, matching the first combined keyword label and each second keyword label one by one, and calculating a matching index of the corresponding second keyword label;
step 5: extracting a third keyword label with the matching index larger than a preset threshold value, and recommending resource information matched with the third keyword label;
the method for identifying and matching the similar keywords of the keywords in the first initial keyword label to obtain a corresponding first similar keyword label, and combining the corresponding first similar keyword label with the first initial keyword label to obtain a first combined keyword label comprises the following steps:
matching similar keywords corresponding to each first keyword in the first initial keyword label from a keyword matching database, so as to obtain a first similar keyword label;
performing label combination on the first initial keyword label and the first similar keyword label to obtain a first combined keyword label;
wherein, based on the first combined keyword label and each second keyword label in the second keyword label database to match one by one, calculating the matching index of the corresponding second keyword label comprises:
matching each second keyword label in the database of the first combined keyword labels and each second keyword label one by one, and determining a keyword matching index of each matching result;
；
wherein,keyword matching indexes corresponding to the matching results; />Refers to the inclusion of the first merged keyword tagThe first merging key word is +.>Words and phrases; />Means +.f contained in the second keyword tag database corresponding to the second keyword tag h>The%>Words and phrases; />、/>The part-of-speech conversion coefficient of the current word;、/>meaning the semantic conversion coefficient of the current word; a1 represents part-of-speech weights; a2 represents semantic weights; nh is the total number of keywords contained in the corresponding second keyword label h.
2. The resource recommendation method based on an industrial chain as claimed in claim 1, wherein obtaining operation information of a user on a current industrial chain when making a resource request as reference resource information, and extracting keywords in the reference information to form a first initial keyword tag, comprises:
acquiring all operation information when a target user makes a resource request in a preset time range on a current industrial chain, and taking the operation information as control resource information;
extracting words from all analysis sentences in the comparison resource information to obtain a first word set of the comparison resource information;
acquiring word characteristics of each word of the first word set and sentence characteristics of each word in a corresponding analysis sentence;
and extracting keywords from the comparison resource information based on the word characteristics and the sentence characteristics on the basis of a machine learning model to obtain all keywords of the comparison resource information, and further obtaining a first initial keyword label.
3. The resource recommendation method based on an industrial chain of claim 2, wherein the extracting keywords of the comparison resource information based on the word features and the sentence features based on the machine learning model to obtain all keywords of the current comparison resource information, and further obtain a first initial keyword tag, includes:
extracting historical keywords in all historical control resource information of the resource management background and the criticality of each historical keyword, and training an initial learning model to obtain a machine learning model;
analyzing word characteristics of each word of a first word set of current comparison resource information and sentence characteristics of corresponding words in corresponding analysis sentences based on the machine learning model, acquiring the criticality of each first word in the first word set, eliminating the same keywords to obtain residual keywords and obtaining first initial keyword labels;
the first initial keyword tag comprises an operation serial number of corresponding operation information and a word sequence number of the rest keywords related to the operation information.
4. The resource recommendation method based on an industrial chain as claimed in claim 1, wherein obtaining resource information to be recommended in a resource management background, and extracting keywords of each resource information to be recommended, and forming a second keyword tag database, comprises:
acquiring all effective resource information of the resource management background to obtain a plurality of resource information to be recommended;
calculating the relevance of each word in each analysis statement of each resource information to be recommended and the rest words in the analysis statement, so as to obtain the weight coefficient of each word in the current analysis statement, and creating a weight graph of each resource information to be recommended according to the weight coefficients of all words in the current analysis statement and the weight coefficient of the current analysis statement;
in the process of recommending historical resource information, the position of the corresponding vertex in the weighted graph is used as the probability of recommending keywords, and is used as the vertex recommending probability of each vertex in the corresponding weighted graph;
sequencing all vertex recommendation probabilities in each weighted graph, extracting second keywords in the corresponding resource information to be recommended, and obtaining corresponding second keyword labels;
a second keyword tag database is constructed based on all of the second keyword tags.
5. The resource recommendation method based on an industrial chain as claimed in claim 1, wherein extracting a third keyword tag having a matching index greater than a preset threshold value and recommending resource information matched with the third keyword tag comprises:
obtaining a matching index of the first combined keyword tag and the current keyword tag of the second keyword tag database, and comparing the matching index with a preset threshold;
if the matching index is larger than a preset threshold value, extracting a second keyword label corresponding to the current matching index in the second keyword label database as a third keyword label;
performing resource recommendation based on the resource information matched with the current third keyword label;
and if the matching index is at the boundary value of the preset matching range, judging whether the current information to be recommended is recommended or not based on the information number of the information to be recommended.
6. The resource recommendation method based on an industrial chain as claimed in claim 5, wherein if the matching index is at a boundary value of a preset matching range, determining whether the current information to be recommended is recommended based on the information number of the information to be recommended includes:
determining the number of the resource information to be recommended, which is obtained by acquiring the boundary between the matching indexes of all the resource information to be recommended and the contrast resource information in the resource management background and is in a preset matching range;
simultaneously, the number of the keywords of the first combined keyword tag corresponding to the comparison resource information is obtained;
if the number of the resource information to be recommended is larger than the preset multiple of the number of the keywords in the first combined keyword tag, judging that the current resource information to be recommended is not recommended;
if the number of the resource information to be recommended is not more than the preset multiple of the number of the keywords of the first combined keyword tag, judging that the current information to be recommended is recommended.
7. The resource recommendation method based on an industrial chain as claimed in claim 6, wherein after extracting a third keyword tag having a matching index greater than a preset threshold value and recommending resource information matched with the third keyword tag, further comprising:
acquiring all third keyword labels, carrying out the sequence from high to low on the matching index of each third keyword label to obtain the arrangement sequence of the corresponding third keyword labels, and carrying out corresponding sequence recommendation on the information to be recommended based on the current sequence;
the information to be recommended corresponding to the third keyword label with the highest matching index is located at the first position of the current industrial chain recommendation position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211555850.9A CN116244496B (en) | 2022-12-06 | 2022-12-06 | Resource recommendation method based on industrial chain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211555850.9A CN116244496B (en) | 2022-12-06 | 2022-12-06 | Resource recommendation method based on industrial chain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116244496A CN116244496A (en) | 2023-06-09 |
| CN116244496B true CN116244496B (en) | 2023-12-01 |
Family
ID=86630258
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211555850.9A Active CN116244496B (en) | 2022-12-06 | 2022-12-06 | Resource recommendation method based on industrial chain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116244496B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106598940A (en) * | 2016-11-01 | 2017-04-26 | 四川用联信息技术有限公司 | Text similarity solution algorithm based on global optimization of keyword quality |
| CN106611059A (en) * | 2016-12-28 | 2017-05-03 | 北京小米移动软件有限公司 | Method and device for recommending multi-media files |
| CN106776881A (en) * | 2016-11-28 | 2017-05-31 | 中国科学院软件研究所 | A kind of realm information commending system and method based on microblog |
| CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
| CN108287916A (en) * | 2018-02-11 | 2018-07-17 | 北京方正阿帕比技术有限公司 | A kind of resource recommendation method |
| CN108334490A (en) * | 2017-04-07 | 2018-07-27 | 腾讯科技（深圳）有限公司 | Keyword extracting method and keyword extracting device |
| CN109637642A (en) * | 2018-12-04 | 2019-04-16 | 平安医疗健康管理股份有限公司 | Medical institutions' dialysis horizontal wire examination method, device, equipment and storage medium |
| CN109815333A (en) * | 2019-01-14 | 2019-05-28 | 金蝶软件（中国）有限公司 | Information acquisition method, device, computer equipment and storage medium |
| CN111104803A (en) * | 2019-12-31 | 2020-05-05 | 科大讯飞股份有限公司 | Semantic understanding processing method, device and equipment and readable storage medium |
| CN111782772A (en) * | 2020-07-24 | 2020-10-16 | 平安银行股份有限公司 | Text automatic generation method, device, equipment and medium based on OCR technology |
| CN111814490A (en) * | 2020-06-05 | 2020-10-23 | 江苏遨信科技有限公司 | Method and system for improving semantic recognition capability through robot self-learning |
| CN114911917A (en) * | 2022-07-13 | 2022-08-16 | 树根互联股份有限公司 | Asset meta-information searching method and device, computer equipment and readable storage medium |
-
2022
- 2022-12-06 CN CN202211555850.9A patent/CN116244496B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106598940A (en) * | 2016-11-01 | 2017-04-26 | 四川用联信息技术有限公司 | Text similarity solution algorithm based on global optimization of keyword quality |
| CN106776881A (en) * | 2016-11-28 | 2017-05-31 | 中国科学院软件研究所 | A kind of realm information commending system and method based on microblog |
| CN106611059A (en) * | 2016-12-28 | 2017-05-03 | 北京小米移动软件有限公司 | Method and device for recommending multi-media files |
| CN108334490A (en) * | 2017-04-07 | 2018-07-27 | 腾讯科技（深圳）有限公司 | Keyword extracting method and keyword extracting device |
| CN108287916A (en) * | 2018-02-11 | 2018-07-17 | 北京方正阿帕比技术有限公司 | A kind of resource recommendation method |
| CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
| CN109637642A (en) * | 2018-12-04 | 2019-04-16 | 平安医疗健康管理股份有限公司 | Medical institutions' dialysis horizontal wire examination method, device, equipment and storage medium |
| CN109815333A (en) * | 2019-01-14 | 2019-05-28 | 金蝶软件（中国）有限公司 | Information acquisition method, device, computer equipment and storage medium |
| CN111104803A (en) * | 2019-12-31 | 2020-05-05 | 科大讯飞股份有限公司 | Semantic understanding processing method, device and equipment and readable storage medium |
| CN111814490A (en) * | 2020-06-05 | 2020-10-23 | 江苏遨信科技有限公司 | Method and system for improving semantic recognition capability through robot self-learning |
| CN111782772A (en) * | 2020-07-24 | 2020-10-16 | 平安银行股份有限公司 | Text automatic generation method, device, equipment and medium based on OCR technology |
| CN114911917A (en) * | 2022-07-13 | 2022-08-16 | 树根互联股份有限公司 | Asset meta-information searching method and device, computer equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116244496A (en) | 2023-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108363790B (en) | Method, device, equipment and storage medium for evaluating comments | |
| CN109829104B (en) | Semantic similarity based pseudo-correlation feedback model information retrieval method and system | |
| US10565533B2 (en) | Systems and methods for similarity and context measures for trademark and service mark analysis and repository searches | |
| WO2021093755A1 (en) | Matching method and apparatus for questions, and reply method and apparatus for questions | |
| CN111125334B (en) | Search question-answering system based on pre-training | |
| CN110727779A (en) | Question-answering method and system based on multi-model fusion | |
| CN111222305B (en) | Information structuring method and device | |
| CN111291188B (en) | Intelligent information extraction method and system | |
| CN112069298A (en) | Human-computer interaction method, device and medium based on semantic web and intention recognition | |
| CN111105209A (en) | Job resume matching method and device suitable for post matching recommendation system | |
| CN112307182B (en) | Question-answering system-based pseudo-correlation feedback extended query method | |
| CN114416942A (en) | Automatic question-answering method based on deep learning | |
| CN113282711B (en) | Internet of vehicles text matching method and device, electronic equipment and storage medium | |
| CN112328800A (en) | System and method for automatically generating programming specification question answers | |
| CN111858888A (en) | Multi-round dialogue system of check-in scene | |
| CN111401065A (en) | Entity identification method, device, equipment and storage medium | |
| CN113569011A (en) | Training method, device and equipment of text matching model and storage medium | |
| CN111241397A (en) | Content recommendation method and device and computing equipment | |
| CN115827819A (en) | Intelligent question and answer processing method and device, electronic equipment and storage medium | |
| CN113297351A (en) | Text data labeling method and device, electronic equipment and storage medium | |
| Celikyilmaz et al. | A graph-based semi-supervised learning for question-answering | |
| CN114491079A (en) | Knowledge graph construction and query method, device, equipment and medium | |
| CN114722198A (en) | Method, system and related device for determining product classification code | |
| CN113535960A (en) | Text classification method, device and equipment | |
| CN113570380A (en) | Service complaint processing method, device and equipment based on semantic analysis and computer readable storage medium |
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 |