CN116610853A - Search recommendation method, search recommendation system, computer device, and storage medium - Google Patents
Search recommendation method, search recommendation system, computer device, and storage medium Download PDFInfo
- Publication number
- CN116610853A CN116610853A CN202210118904.9A CN202210118904A CN116610853A CN 116610853 A CN116610853 A CN 116610853A CN 202210118904 A CN202210118904 A CN 202210118904A CN 116610853 A CN116610853 A CN 116610853A
- Authority
- CN
- China
- Prior art keywords
- search
- user
- words
- preference
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000006399 behavior Effects 0.000 claims abstract description 14
- 230000011218 segmentation Effects 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 description 10
- 238000004458 analytical method Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000007723 transport mechanism 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
- 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
- G06F16/906—Clustering; Classification
-
- 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
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the application provides a search recommendation method, a search recommendation system, computer equipment and a storage medium, and belongs to the technical field of search. The method comprises the following steps: acquiring index words corresponding to search sentences input by a user; acquiring search recommended words of the user, wherein the search recommended words are used for associating historical search behaviors of the user; and obtaining search results according to the index words and the search recommended words. According to the technical scheme, the search results are improved by associating the historical search preference of the user during search, and the sorting of the search results is optimized, so that the search intention of the user can be judged more accurately, and the user experience is improved.
Description
Technical Field
The present application relates to the field of search technologies, and in particular, to a search recommendation method, a search recommendation system, a computer device, and a storage medium.
Background
With the continuous development of the telecommunication industry, service knowledge is continuously updated, and telecommunication customer service personnel search through a unified knowledge management system frequently in order to acquire relevant service information, and the same type of knowledge can be searched frequently according to factors such as working positions, habits and the like. Accurate and efficient service information retrieval is very important to recommend target articles which are most in line with search expectations to search staff.
The current search ranking method is as follows: after the user inputs the keywords as search words, the system background server directly matches the words to be searched with the search words in the knowledge base to obtain a plurality of search target articles. Then, for each article, a ranking score for each article is calculated according to the "relevance score", and the articles are ranked and recommended to the user according to the score. Wherein the relevance score characterizes the degree of agreement of the user's search term with the article, calculated by the system server using a specific algorithm. However, when some search sentences of the user are not matched with the search keywords, the search intention of the user cannot be accurately judged, so that the search result cannot be satisfied with the user.
Disclosure of Invention
The embodiment of the application mainly aims to provide a search recommendation method, a search recommendation system, computer equipment and a storage medium, and aims to solve the problem that the existing knowledge search cannot be associated with the historical behavior of a user, the search intention of the user cannot be accurately judged, so that the accuracy of a search result is insufficient, and the experience of the user is improved.
In a first aspect, an embodiment of the present application provides a search recommendation method, including:
acquiring index words corresponding to search sentences input by a user; acquiring search recommended words of the user, wherein the search recommended words are used for associating historical search behaviors of the user; and obtaining search results according to the index words and the search recommended words.
In a second aspect, an embodiment of the present application further provides a search recommendation system, including:
the index word acquisition module is used for acquiring index words corresponding to search sentences input by a user; the user preference module is used for acquiring search recommended words of the user, wherein the search recommended words are used for associating historical search behaviors of the user; and the search module is used for acquiring search results according to the index words and the search recommended words.
In a third aspect, embodiments of the present application further provide a computer device comprising a processor, a memory, a computer program stored on the memory and executable by the processor, and a data bus for enabling a connected communication between the processor and the memory, wherein the computer program, when executed by the processor, implements the steps of any of the search recommendation methods as provided in the present specification.
In a fourth aspect, embodiments of the present application also provide a storage medium for computer readable storage storing one or more programs executable by one or more processors to implement the steps of any of the methods of searching for recommendations as provided in the present specification.
The embodiment of the application provides a search recommendation method, a search recommendation system, computer equipment and a storage medium. On the other hand, the search recommended words representing the search preference of the user and the preference category are used together with the index words and the classification information contained in the search statement to carry out relevance scoring on the articles and sort the articles according to the scoring result, so that the articles which are more in line with the search intention of the user can be displayed in front. Further, by deep learning of the user search and review data, the search preferences of the user are predicted and analyzed, and the analysis results are used to correct the search preferences of the user, so that the search preferences of the user can be intelligently adjusted according to the search history. The search recommendation result which better accords with the intention of the user is obtained, and the satisfaction degree of the user is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of a search recommendation method according to an embodiment of the present application;
FIG. 2 is a flow chart of a search recommendation method in the telecommunications industry according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a dictionary tree constructed in accordance with one embodiment of the present application;
FIG. 4 is a schematic flow chart of a search recommendation system according to an embodiment of the present application;
fig. 5 is a schematic block diagram of a computer device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The flow diagrams depicted in the figures are merely illustrative and not necessarily all of the elements and operations/steps are included or performed in the order described. For example, some operations/steps may be further divided, combined, or partially combined, so that the order of actual execution may be changed according to actual situations.
It is to be understood that the terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The embodiment of the application provides a search recommendation method, a search recommendation system, computer equipment and a storage medium. The search recommendation method can be applied to mobile terminals, and the mobile terminals can be mobile phones, tablet computers, notebook computers, desktop computers, personal digital assistants, wearable devices and other electronic devices.
Some embodiments of the application are described in detail below with reference to the accompanying drawings. The following embodiments and features of the embodiments may be combined with each other without conflict.
Referring to fig. 1, fig. 1 is a flowchart of a search recommendation method according to an embodiment of the present application.
As shown in fig. 1, the search recommendation method includes steps S101 to S103.
Step S101, obtaining index words corresponding to search sentences input by a user;
specifically, word segmentation matching is performed on search sentences input by a user and a preset dictionary tree, so that index words corresponding to the search sentences are obtained. Firstly, constructing a preset dictionary tree, classifying articles in a preset knowledge base, and constructing the preset dictionary tree according to the word segmentation result of the contents of the articles and the category to which the articles belong. Further, the service category corresponding to the index word can be obtained according to the category in the preset dictionary tree.
For convenience of explanation, the description will be given by taking the establishment of the classification knowledge index in the specific scene of the telecommunications industry as an example, and it should be noted that the application is also applicable to other business scenes.
Specifically, firstly collecting professional terms and articles in the telecommunication industry, and classifying the articles according to the service; adding technical terms and common terms in the telecommunication industry into a word stock, and respectively carrying out coarse-granularity word segmentation and fine-granularity word segmentation on the content of the knowledge articles and removing duplication to obtain word segmentation results, wherein the coarse-granularity word segmentation is carried out according to the custom words when sentences contain the custom words in the telecommunication industry. And constructing knowledge base indexes of different types according to article classification, constructing a dictionary tree, and storing the word segmentation result and article classification information into the dictionary tree. And inputting the search sentences of the user into a preset dictionary tree for matching, and obtaining index words matched with the search sentences.
Illustratively, knowledge articles may be segmented by an ANSJ segmenter.
Step S102, acquiring search recommended words of a user, wherein the search recommended words are used for associating historical search behaviors of the user;
specifically, a search recommended word preset by a user can be obtained from a user database according to the user ID of the user, and the search recommended word is used for searching together with an index word matched with a search sentence. And furthermore, the historical search behaviors of the user can be associated in each search, on one hand, the input information of the search is expanded, and on the other hand, the search result is more similar to the search intention of the user based on more search recommended words of the user.
Further, in order to better predict and analyze the search preference of the user, the historical search of the user and the reading behavior of the articles can be analyzed based on the neural network, and then the search recommended words of the user are updated and maintained according to the analysis result. Specifically, through collecting information of a user and data of searching and reading, searching preference of the user is obtained by analyzing the information of the user and searching history based on a preset BP neural network model, and searching recommended words of the user are updated according to the searching preference.
And step S103, obtaining search results according to the index words and the search recommended words.
And searching articles in the knowledge base according to the index words and the search recommended words, and specifically, obtaining required search results by filtering articles which do not contain the index words and the search recommended words in the knowledge base.
Further, in order to further expand the keyword for searching, the synonym related to the index word can be searched from a preset relational database, and a matched searching result is obtained from a preset knowledge base according to the index word, the synonym and the searching recommended word.
In order to better promote the satisfaction degree of the search results to the user, the search results can be ranked, so that the search results which are more relevant to the search intention of the user are displayed in front, and the user can quickly locate the required knowledge articles.
Specifically, the relevance scoring is carried out on the search results, and the search results are ranked according to the obtained relevance score to obtain recommended results.
Since the search results need to be ranked in order to rank articles that more satisfy the user's search intent in a more advanced position. The search results need to be scored in relevance, and articles with higher scores are closer to the search intention of the user according to the scoring results, so that the search results are ranked in a high-to-low order to obtain ranked search results, namely recommended results, and the ranked search results are displayed to the user.
Further, in order to better embody the importance of the search vocabulary in each search, the relevance scoring is performed on the search results by setting different weights for the index words, the synonyms and the search recommended words. Specifically, the index word is a search entity word matched according to the search sentence of the user, so that the highest weight is set, the synonym of the index word is set to the second highest weight, and the search recommended word of the user represents the historical search behavior of the user, and the relationship with the current search is possibly not big, so that the third highest weight is set. And setting a strategy according to the weights of index words, synonyms and search recommended words, scoring the relevance of the articles in all search results according to the three words and the weights of the words, and sorting the articles according to the relevance scores.
For example, the calculation of relevance scores may be performed on articles in search results based on TF-IDF.
Further, optimizing the recommendation result according to the preference classification to obtain an optimized recommendation result; wherein optimizing the recommendation result according to the preference classification comprises: and weighting the relevance scores of the search results belonging to the preference classification, and sorting the search results according to the obtained weighted relevance scores to obtain optimized recommendation results.
The preference classification of the user can be set by the user in advance according to the service requirement of the user, and in order to maintain the preference classification of the user more intelligently, the historical behavior and the dimensional information of the user can be intelligently analyzed based on the neural network, so that the preference classification condition of the user can be intelligently maintained and updated.
Illustratively, collecting information of the user includes: collecting information of each dimension of a user, such as user post information, client type facing and working period; the search history data of the user includes: the browsing sequence of the articles, the times of reading the articles, the time of reading the articles, whether the articles are collected or not and the like of the user are taken as neurons of a neural network input layer.
Establishing a BP neural network model for each category, taking the information data of each dimension as the input of the BP neural network, and calculating and outputting the preference scores of the most recently read articles of the user in the most recent time period by using an excitation function in an implicit layer, wherein the calculation formula is as follows:
is set in the above-described range.
Therefore, the preference scores of the recently read articles of the user are obtained according to the BP neural network structure, articles with high preference scores of the recently read articles are counted, the preference classification scores of the user are calculated according to the self-defined algorithm according to the types, the preference scores and the classified article number of the articles, and the types with high scores are used as the preference classifications of the user.
And counting the articles with high preference scores of the most recently read articles, analyzing coarse-granularity words with high frequency of the text according to the articles, and correcting the words serving as search recommended words of the user to search recommended words in a user database.
Through deep learning of user searching and reading data, the search preference of the user is predicted and analyzed, and the analysis result is used for correcting the search preference of the user, so that the search preference of the user can be intelligently adjusted according to the search history.
Further, according to the historical search of the user, the access frequency of the user can be recorded for the articles ranked in front, for example, the access frequency of the users of the articles is little or the users do not access, the articles are not considered to be in accordance with the search intention of the users, and the articles are subjected to the reduction operation in the sorting process.
Further, category analysis can be performed on articles with low user access frequency, dislike categories of the users can be identified, and the dislike categories are used as basis for reducing scores when the search results are ranked.
For example, for the record of the access frequency of the first 100 articles recommended to the user, the articles are associated with the user ID through the ID of the articles in the database, the access frequency is counted, the access frequency is obtained in the sorting process, and when the access frequency is small or 0, the corresponding articles are reduced.
For example, for the articles with small access times or 0, the categories to which the low-frequency access articles belong are obtained, when the low-frequency access articles in a certain category exceed a certain number, namely the category is considered as a non-preference category of the user, and when the search results are ranked, the search results belonging to the non-preference category are subtracted.
And sorting the articles in the search results according to the score from high to low, and returning the sorted search results to the user as recommended results. By using the search recommended words representing the search preference of the user and the preference categories together with the search index words to score the relevance of the articles and sort the articles according to the scoring results, the articles which are more in line with the search intention of the user can be displayed in front, and the user can find the required articles more quickly.
According to the search recommendation method provided by the embodiment of the application, the index words contained in the search sentences are extracted based on the dictionary tree of the knowledge base, the search preference of the user is obtained for searching, the search sentences are expanded, and more results conforming to the search intention of the user are obtained. On the other hand, search recommended words representing the search preference of the user and preference categories are used together with search index words to score the relevance of the articles and rank the articles according to the scoring results, so that the articles more conforming to the search intention of the user are displayed in front. Further, by deep learning of the user search and review data, the search preferences of the user are predicted and analyzed, and the analysis results are used to correct the search preferences of the user, so that the search preferences of the user can be intelligently adjusted according to the search history. The search recommendation result which better accords with the intention of the user is obtained, and the satisfaction degree of the user is improved.
In addition, the embodiment of the application also provides two specific embodiments of the search recommendation method, which are specifically as follows:
example 1
Referring to fig. 2 and 3, fig. 2 is a flow chart of a search recommendation method in the telecommunication industry according to an embodiment of the application, and fig. 3 is a diagram of a dictionary tree constructed according to an embodiment of the application. The embodiment designates search recommendations in a specific scenario for simplicity only, and it should be noted that the present application is also applicable to search recommendations in other scenarios.
Step 1, collecting a telecommunications industry knowledge article, for example: "internet national data transfer", "internet data center", "call center". And categorizing the articles by content: the Internet national data transmission belongs to basic telecommunication service, the Internet data center and the call center belong to value-added telecommunication service.
Step 2, word segmentation is carried out on the knowledge articles, and word segmentation results of 'domestic data transmission in the Internet' comprise: the word segmentation results of the Internet, domestic, data and transmission include: "Internet", "data", "center", "data center", and "call center" word segmentation results include: "call", "center". The results of the word segmentation for the above knowledge articles are shown in the following table:
and constructing a dictionary tree shown in fig. 3 according to the word segmentation result and classification shown in the table.
Step 3, obtaining the user ID (such as Test 1), and obtaining the search statement input by the user, such as 'Internet data'.
And 4, matching the search sentences of the user through dictionary tree nodes to obtain index words of 'Internet', 'data', and matched classifications of 'basic telecom service', 'value added telecom service'.
And 5, acquiring search recommended words, such as 'domestic', in the user search preferences stored in the database through the user ID.
And 6, carrying out matching search on different classification indexes according to the classification to which the index word and the search recommended word belong, and obtaining matched articles comprises the following steps: "internet national data transfer", "internet data center".
Step 7, setting different weight values for the search word and the recommended word, wherein the weight of the index word is larger than that of the search recommended word, grading the matched articles by using a TF-IDF grading algorithm according to the weights of the search word and the search recommended word, and calculating the word frequency TF of the word w in the document d, namely the ratio of the occurrence times count (w, d) of the word w in the document d to the total word number size (d) in the document d: tf (w, d) =count (w, d)/size (d).
"Internet national data transfer", "Internet data center", "Call center" statistics are as follows:
the inverse document frequency idf of the word w in the whole document set, i.e., the logarithm idf=log (n/docs (w, D)) of the ratio of the total number of documents n to the number of files docs (w, D) where the word w appears.
For example, when the total number of documents is 10000, the following data can be counted:
| tf(w,d) | internet network | Data | Domestic use |
| Number of occurrences | 200 | 400 | 2000 |
| idf | Log(10000/200) | Log(10000/400) | Log(10000/2000) |
The Tf-idf model calculates a value for each document d and a query string q consisting of keywords w [1]. W [ k ] based on Tf and idf, for representing the matching degree of the query string q to the document d:
tf-idf(q,d)＝sum{i＝1..k|tf-idf(w[i],d)}＝sum{i＝1..k|tf(w[i],d)*idf(w[i])}
further, a weighting coefficient H is added, and the final query matching degree is:
tf-idf(q,d,H)＝sum{i＝1..k|tf-idf(w[i],d)}＝sum{i＝1..k|H*tf(w[i],d)*idf(w[i])}
for example, the predetermined preference score ratio is as follows:
| index words | Searching for recommended words | Preference classification | Disfavor classification | |
| Weighting coefficient H | 1 | 0.2 | 1.2 | 0.8 |
The scores calculated are as follows:
based on the correlation calculations described above, the "Internet national data transfer" score is greater than the "Internet data center".
Further, the database is queried through the user ID to obtain a preference classification such as basic telecom service, and then articles belonging to the preference classification in the search result are processed according to the preference classification of basic telecom service, and further the Internet domestic data transmission is further classified.
And 8, finally, sorting the search results according to the scores from large to small to form a search time result list, and finally, returning the search result list to the user.
Example two
The second embodiment of the application provides application of the user preference screening method in a telecommunication customer service system. In this embodiment, for simplicity, the user preference screening in a specific scenario is only needed, and it should be noted that the present application is also applicable to preference screening in other application scenarios.
Step 1, collecting information of each dimension of a user, such as user post information (business consultation, expense inquiry, business handling, new business promotion and the like), client-oriented type (home client, administrative enterprise client, public telephone, wireless local telephone) and working period. And search and browse data of articles by a user, including: browsing sequence, number of reading articles, time of reading articles, collection or not, etc., and each dimension information of the user and the search browsing data are used as neurons of an input layer.
And 2, establishing a BP neural network model for each classification, taking the information data of each dimension as the input of the BP neural network, calculating and outputting the preference scores of the reading articles in the latest time period by using an excitation function in an implicit layer, and acquiring the preference scores of the latest reading articles of the user according to the BP neural network structure.
And step 3, counting the articles with high preference scores of the recently read articles, calculating the preference classification scores of the users according to the classification, preference scores and the quantity of classified articles and a user-defined algorithm, and classifying the articles with high scores as the preference of the users.
And 4, counting the articles with high preference scores of the most recently read articles, analyzing coarse-grained words with the same text and higher frequency according to the articles, and correcting and updating the search recommended words in the user database by taking the words as the user search recommended words.
And 5, continuously collecting data of the reading behaviors of the user by using the sequencing treaty pushed to the user, wherein the method comprises the following steps of: the click times, the reading time length, whether the article is collected or not and the like of the matching article are used as the input of the step 1. The browsing sequence is forward, and the user is reflected to like an article by clicking for many times, reading time is long and collecting the article; otherwise, the user is stated to dislike the article.
Finally, the search recommendation words and preference classifications of the user are updated by historical behavior of the user and user information based on the BP neural network.
Referring to fig. 4, fig. 4 is a schematic view of a scenario in which the search recommendation system provided in this embodiment is implemented, and as shown in fig. 4, the search recommendation system provided in the foregoing embodiment includes: an index word acquisition module 201, a user preference acquisition module 202, and a search module 203.
Wherein, the index word obtaining module 201 includes: a search request acquisition module 2011 and an index word matching module 2012. The search request acquisition module 2011 is configured to receive a search sentence of a user, and the index word matching module 2012 is configured to extract an index word and a category in the input search sentence. The specific method for generating the index words and the classifications comprises the following steps: and extracting index words contained in the search sentences and classification information corresponding to the index words from the dictionary tree through matching the search sentences with a preset dictionary tree.
The user preference obtaining module 202 includes a search recommendation word obtaining module 2021 and a preference classification obtaining module 2022, where the search recommendation word obtaining module 2021 is configured to obtain a search recommendation word of a user from a database according to ID information of the user; the preference category acquiring module 2033 is configured to acquire a preference category of a user from a database according to ID information of the user, and use the preference category as one of the bases for ranking search results subsequently.
The search module 203 includes a synonym acquisition module 2031 and a matching search module 2032. The synonym obtaining module 2031 is configured to obtain synonyms of the index words extracted from the search statement; the matching search module 2032 is configured to perform search matching from the knowledge database according to the index word, the synonym, and the search recommendation word of the user to obtain a search result.
Further, the search recommendation system further includes: the sorting module 204 may be specifically divided into: a weight labeling module 2041, a relevance scoring module 2042, and a ranking recommendation module 2043. The weight marking module 2041 is used for giving different weights to the index words, the synonyms and the search recommended words of the user; the relevance scoring module 2042 is configured to score relevance of the search results according to the index word, the synonym word, the search recommendation word, and the corresponding different weights, and the scoring results are used as a basis for ranking the search results. The ranking recommendation module 2043 is configured to rank the search results according to the relevance score result of the search results and the preference classification of the user in order of the score from high to low to form a recommendation result, and return the searched recommendation result to the user.
Further, the search recommendation system further includes: the user preference filtering module 205 specifically includes: a user information acquisition module 2051, a search history acquisition module 2052, a user preference analysis module 2053, and a user preference correction module 2054. The user information acquiring module 2051 is configured to collect basic information of a user; the browsing history obtaining module 2052 is configured to obtain history data (such as an order of browsing articles, a number of reading, and whether to collect) of articles searched for and reviewed by the user; the user preference analysis module 2053 is configured to predict a preference degree of a user for an article according to information of the user and a browsing history of the article in combination with a BP neural network, extract articles with high preference degree, collect and fuse the number of the preference articles and the preference degree, extract preference classifications, and screen out search recommended words according to the preference articles according to a predetermined screening principle; the user preference correction module 2054 is configured to update preference categories and search recommendations in the user database based on analysis and prediction results of the BP neural network.
Further, the search recommendation system further includes: the knowledge base classification index module 206 can be specifically divided into: a knowledge base classification module 2061, a knowledge base word segmentation module 2062, and a dictionary tree construction module 2063. The knowledge base classification module 2061 is used for classifying articles of the knowledge base according to the service; the knowledge base word segmentation module 2062 is used for extracting professional vocabulary and other commonly used custom vocabulary in the knowledge base articles as index words; the dictionary tree construction submodule 2063 is used for constructing a dictionary tree according to the category to which the index word article belongs.
Referring to fig. 5, fig. 5 is a schematic block diagram of a computer device according to an embodiment of the present application.
As shown in fig. 5, the computer device 300 includes a processor 301 and a memory 302, the processor 301 and the memory 302 being connected by a bus 303, such as I 2 C (Inter-integrated Circuit) bus.
In particular, the processor 301 is used to provide computing and control capabilities, supporting the operation of the overall computer device. The processor 301 may be a central processing unit (Central Processing Unit, CPU), the processor 301 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field-programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. Wherein the general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
Specifically, the Memory 302 may be a Flash chip, a Read-Only Memory (ROM) disk, an optical disk, a U-disk, a removable hard disk, or the like.
It will be appreciated by those skilled in the art that the architecture shown in fig. 5 is merely a block diagram of a portion of the architecture in connection with an embodiment of the present application and is not intended to limit the computer device to which an embodiment of the present application may be applied, and that a particular server may include more or fewer components than shown, or may combine certain components, or may have a different arrangement of components.
The processor is used for running a computer program stored in the memory and realizing any one of the search recommendation methods provided by the embodiment of the application when the computer program is executed.
In an embodiment, the processor is configured to run a computer program stored in the memory and to implement the following steps when the computer program is executed:
acquiring index words corresponding to search sentences input by a user;
acquiring search recommended words of a user, wherein the search recommended words are used for associating historical search behaviors of the user;
and obtaining search results according to the index words and the search recommended words.
In an embodiment, the processor, when implementing the search recommendation method, is configured to implement: and carrying out relevance scoring on the search results, and sorting the search results according to the obtained relevance score to obtain recommended results.
In an embodiment, the processor, when implementing the search recommendation method, is configured to implement: optimizing the recommendation result according to the preference classification to obtain an optimized recommendation result; wherein optimizing the recommendation result according to the preference classification comprises: and weighting the relevance scores of the search results belonging to the preference classification, and sorting the search results according to the obtained weighted relevance scores to obtain optimized recommendation results.
In an embodiment, the processor, when implementing the search recommendation method, is configured to implement: collecting information and search history of a user, and analyzing the information and the search history of the user based on a preset BP neural network model to obtain search preference of the user; the search recommendation words and the preference classifications of the users are updated based on the search preferences.
In an embodiment, when implementing obtaining an index word corresponding to a search term input by a user, the processor is configured to implement: and carrying out word segmentation matching on the search sentences input by the user and a preset dictionary tree to obtain index words corresponding to the search sentences.
In an embodiment, the processor, when implementing the search recommendation method, is configured to implement: classifying articles in a preset knowledge base, and constructing a preset dictionary tree according to the word segmentation result of the content of the articles and the category to which the articles belong.
In one embodiment, the processor, when implementing obtaining the search results based on the index word and the search recommendation word, is configured to implement: searching synonyms associated with the index words from a preset relational database; and obtaining search results according to the index words, the synonyms and the search recommended words.
In an embodiment, the processor, when implementing relevance scoring of the search results, is to implement: and setting different weights for the index words, the synonyms and the search recommended words to score the relevance of the search results, so as to obtain the relevance score of the search results.
It should be noted that, for convenience and brevity of description, specific working processes of the above-described computer device may refer to corresponding processes in the foregoing search recommendation method embodiments, which are not described herein again.
The embodiment of the application also provides a storage medium for computer readable storage, the storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement the steps of any search recommendation method as provided in the embodiment of the application.
The storage medium may be an internal storage unit of the computer device of the foregoing embodiment, for example, a hard disk or a memory of the computer device. The storage medium may also be an external storage device of a computer device, such as a plug-in hard disk provided on the computer device, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), or the like.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware embodiment, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
It should be understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments. While the application has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions may be made therein without departing from the spirit and scope of the application as defined by the appended claims. Therefore, the protection scope of the application is subject to the protection scope of the claims.
Claims (11)
1. A search recommendation method, comprising:
acquiring index words corresponding to search sentences input by a user;
acquiring search recommended words of the user, wherein the search recommended words are used for associating historical search behaviors of the user;
and obtaining search results according to the index words and the search recommended words.
2. The method of claim 1, wherein after the search results are obtained from the index word and the search recommended word, the method further comprises:
and carrying out relevance scoring on the search results, and sorting the search results according to the obtained relevance scores to obtain recommended results.
3. The search recommendation method according to claim 2, further comprising:
obtaining preference classification of the user, optimizing the recommendation result according to the preference classification, and obtaining an optimized recommendation result;
wherein optimizing the recommendation result according to the preference classification comprises:
and weighting the relevance scores of the search results belonging to the preference classification, and sorting the search results according to the obtained weighted relevance scores to obtain optimized recommendation results.
4. The search recommendation method of claim 3, further comprising:
collecting information and search history of the user, and analyzing the information and the search history of the user based on a preset BP neural network model to obtain search preference of the user;
and updating search recommended words and preference classification of the user based on the search preference.
5. The method for recommending search according to claim 1, wherein the obtaining the index word corresponding to the search term input by the user comprises:
and carrying out word segmentation matching on the search sentences input by the user and a preset dictionary tree to obtain index words corresponding to the search sentences.
6. The search recommendation method of claim 5, further comprising:
classifying articles in a preset knowledge base, and constructing the preset dictionary tree according to the word segmentation result of the content of the articles and the category to which the articles belong.
7. The method of claim 2, wherein the obtaining search results from the index word and the search recommended word further comprises:
searching synonyms associated with the index words from a preset relational database;
and obtaining search results according to the index words, the synonyms and the search recommended words.
8. The search recommendation method of claim 7, wherein said scoring said search results comprises:
and setting different weights for the index words, the synonyms and the search recommended words to score the relevance of the search results, so as to obtain the relevance score of the search results.
9. A search recommendation system, comprising:
the index word acquisition module is used for acquiring index words corresponding to search sentences input by a user;
the user preference module is used for acquiring search recommended words of the user, wherein the search recommended words are used for associating historical search behaviors of the user;
and the search module is used for acquiring search results according to the index words and the search recommended words.
10. A computer device, characterized in that it comprises a processor, a memory, a computer program stored on the memory and executable by the processor, and a data bus for enabling a connected communication between the processor and the memory, wherein the computer program, when being executed by the processor, implements the steps of the search recommendation method according to any one of claims 1 to 8.
11. A storage medium for computer-readable storage, wherein the storage medium stores one or more programs executable by one or more processors to implement the steps of the search recommendation method of any one of claims 1 to 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210118904.9A CN116610853A (en) | 2022-02-08 | 2022-02-08 | Search recommendation method, search recommendation system, computer device, and storage medium |
| PCT/CN2023/074947 WO2023151576A1 (en) | 2022-02-08 | 2023-02-08 | Search recommendation method, search recommendation system, computer device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210118904.9A CN116610853A (en) | 2022-02-08 | 2022-02-08 | Search recommendation method, search recommendation system, computer device, and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116610853A true CN116610853A (en) | 2023-08-18 |
Family
ID=87563626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210118904.9A Pending CN116610853A (en) | 2022-02-08 | 2022-02-08 | Search recommendation method, search recommendation system, computer device, and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN116610853A (en) |
| WO (1) | WO2023151576A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117391824A (en) * | 2023-12-11 | 2024-01-12 | 深圳须弥云图空间科技有限公司 | Method and device for recommending articles based on large language model and search engine |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188186A (en) * | 2019-04-24 | 2019-08-30 | 平安科技（深圳）有限公司 | Content recommendation method, electronic device, equipment and the storage medium of medical field |
| KR102237319B1 (en) * | 2019-07-22 | 2021-04-07 | 주식회사 앱컴파니 | An artificial intelligence system providing customized goods |
| CN111737574B (en) * | 2020-06-19 | 2024-01-26 | 口口相传(北京)网络技术有限公司 | Search information acquisition method, apparatus, computer device and readable storage medium |
| CN113282832A (en) * | 2021-06-10 | 2021-08-20 | 北京爱奇艺科技有限公司 | Search information recommendation method and device, electronic equipment and storage medium |
| CN113343091A (en) * | 2021-06-22 | 2021-09-03 | 力合科创集团有限公司 | Industrial and enterprise oriented science and technology service recommendation calculation method, medium and program |
-
2022
- 2022-02-08 CN CN202210118904.9A patent/CN116610853A/en active Pending
-
2023
- 2023-02-08 WO PCT/CN2023/074947 patent/WO2023151576A1/en unknown
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117391824A (en) * | 2023-12-11 | 2024-01-12 | 深圳须弥云图空间科技有限公司 | Method and device for recommending articles based on large language model and search engine |
| CN117391824B (en) * | 2023-12-11 | 2024-04-12 | 深圳须弥云图空间科技有限公司 | Method and device for recommending articles based on large language model and search engine |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023151576A1 (en) | 2023-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2019214245A1 (en) | Information pushing method and apparatus, and terminal device and storage medium | |
| US11663254B2 (en) | System and engine for seeded clustering of news events | |
| US10146862B2 (en) | Context-based metadata generation and automatic annotation of electronic media in a computer network | |
| CN109885773B (en) | Personalized article recommendation method, system, medium and equipment | |
| WO2021098648A1 (en) | Text recommendation method, apparatus and device, and medium | |
| US8380697B2 (en) | Search and retrieval methods and systems of short messages utilizing messaging context and keyword frequency | |
| CN111008265B (en) | Enterprise information searching method and device | |
| US20160034514A1 (en) | Providing search results based on an identified user interest and relevance matching | |
| CN107180093B (en) | Information searching method and device and timeliness query word identification method and device | |
| US20140214835A1 (en) | System and method for automatically classifying documents | |
| US20080104113A1 (en) | Uniform resource locator scoring for targeted web crawling | |
| CN112148889A (en) | Recommendation list generation method and device | |
| CN104834651B (en) | Method and device for providing high-frequency question answers | |
| Im et al. | Linked tag: image annotation using semantic relationships between image tags | |
| US10747759B2 (en) | System and method for conducting a textual data search | |
| CN110390094B (en) | Method, electronic device and computer program product for classifying documents | |
| WO2018171295A1 (en) | Method and apparatus for tagging article, terminal, and computer readable storage medium | |
| CA2956627A1 (en) | System and engine for seeded clustering of news events | |
| US20120059786A1 (en) | Method and an apparatus for matching data network resources | |
| WO2023151576A1 (en) | Search recommendation method, search recommendation system, computer device and storage medium | |
| CN105159898A (en) | Searching method and searching device | |
| CN108509449B (en) | Information processing method and server | |
| CA3072192A1 (en) | Diversity evaluation in genealogy search | |
| CN114168733B (en) | Rule retrieval method and system based on complex network | |
| CN113656575B (en) | Training data generation method and device, electronic equipment and readable medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |