CN116644172A - Knowledge graph-based operation path recommendation method and related products - Google Patents

Knowledge graph-based operation path recommendation method and related products Download PDF

Info

Publication number
CN116644172A
CN116644172A CN202310629604.1A CN202310629604A CN116644172A CN 116644172 A CN116644172 A CN 116644172A CN 202310629604 A CN202310629604 A CN 202310629604A CN 116644172 A CN116644172 A CN 116644172A
Authority
CN
China
Prior art keywords
operation path
learning model
category
identification information
deep learning
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
Application number
CN202310629604.1A
Other languages
Chinese (zh)
Inventor
张泰铭
成时馨
李玉珊
晋乐乐
闫雪
戴震
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Bank of China Ltd
Original Assignee
Bank of China Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Bank of China Ltd filed Critical Bank of China Ltd
Priority to CN202310629604.1A priority Critical patent/CN116644172A/en
Publication of CN116644172A publication Critical patent/CN116644172A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/335—Filtering based on additional data, e.g. user or group profiles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
    • G06F16/367—Ontology
    • 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
    • G06F40/295—Named entity recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/01—Customer relationship services
    • 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)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Databases & Information Systems (AREA)
  • Business, Economics & Management (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Animal Behavior & Ethology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Finance (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The application provides an operation path recommending method based on a knowledge graph and a related product, which can be applied to the field of artificial intelligence or finance and comprises the following steps: acquiring identification information, category information and an operation path in a target product, wherein the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence; constructing a knowledge graph by utilizing the identification information, the category information and the operation path; based on the knowledge graph, constructing a deep learning model, wherein the deep learning model is used for learning the relevance of the content in the knowledge graph; and inputting the user input corpus into the deep learning model, and outputting a recommended operation path. Therefore, through the knowledge graph and the deep learning model, the corresponding relation between the complete operation path and the operation related description is learned, the corpus can be directly input by the user, the corresponding operation path is matched, the acquisition efficiency of the operation path is improved, and the use experience of the user is improved.

Description

Knowledge graph-based operation path recommendation method and related products
Technical Field
The application relates to the technical field of artificial intelligence or finance, in particular to an operation path recommending method based on a knowledge graph and a related product.
Background
With the development of internet financial products, more functions or programs may be stored in the same financial product. With the increasing of functions, the seat personnel are more faced with storing knowledge documents and classifying knowledge when using an internal knowledge base.
In this case, it is difficult for a user of the financial product or the knowledge base to find a corresponding path of a desired operation in the face of complicated contents. If the user is a customer, the customer viscosity is low, and the product conversion rate is low; if the user is a seat person, the seat person can hardly find the answering problem, the waiting time of the customer is long, and the answering rate of the seat is reduced.
Therefore, how to improve the efficiency of finding the operation path is a technical problem that needs to be solved by the skilled person.
Disclosure of Invention
In view of the above, the embodiment of the application provides an operation path recommending method based on a knowledge graph and related products, aiming at improving the efficiency of finding a needed operation path for a user.
In a first aspect, an embodiment of the present application provides a method for recommending an operation path based on a knowledge graph, including:
acquiring identification information, category information and an operation path in a target product, wherein the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence;
constructing a knowledge graph by utilizing the identification information, the category information and the operation path;
based on the knowledge graph, constructing a deep learning model, wherein the deep learning model is used for learning the relevance of the content in the knowledge graph;
and inputting the user input corpus into the deep learning model, and outputting a recommended operation path.
Optionally, the method constructs a knowledge graph by using the text information, the category information and the operation path,
comprising the following steps:
generating a temporary storage database by using the text information, the category information and the operation path;
performing data cleaning on the temporary storage database to obtain an atomic level text, wherein the atomic level text comprises a document title field, a category field and an operation path field;
and constructing a knowledge graph by using the atomic level text, wherein the knowledge graph takes the category field as an index.
Optionally, the constructing a deep learning model based on the knowledge graph includes:
performing deep learning of named entity recognition on the knowledge graph to respectively obtain the identification information, the category information and the entity type of the operation path;
disambiguating and integrating the entity types to obtain the identification information, the category information and the common category of the operation path;
based on the identification information, the category information and the common category of the operation path, establishing a corresponding relation of the identification information, the category information and the operation path;
and generating a deep learning model by using the corresponding relation among the identification information, the category information and the operation path.
Optionally, the method further comprises:
acquiring an operation path continuously accessed by a user;
training the deep learning model based on the operation paths continuously accessed by the user to obtain an iterative learning model, wherein the iterative learning model is used for learning the relevance of the operation paths continuously accessed by the user;
and responding to the access of a user to a first operation path, inputting the first operation path into the iterative deep learning model, and outputting a linkage operation path.
Optionally, after the inputting the user input corpus into the deep learning model and outputting the recommended operation path, the method further includes:
and responding to the recommended operation path as a recommended document, and extracting and outputting key language segments of the recommended document.
Optionally, the obtaining the identification information, the category information and the operation path in the target product includes:
and capturing identification information, category information and an operation path in the target product by utilizing the web crawler.
In a second aspect, an embodiment of the present application provides an operation path recommendation device based on a knowledge graph, including:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring identification information, category information and an operation path in a target product, the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence;
the map construction module is used for constructing a knowledge map by utilizing the identification information, the category information and the operation path;
the model construction module is used for constructing a deep learning model based on the knowledge graph, and the deep learning model is used for learning the relevance of the content in the knowledge graph;
and the path recommending module is used for inputting the corpus input by the user into the deep learning model and outputting a recommended operation path.
Optionally, the map construction module includes:
the database construction unit is used for generating a temporary storage database by utilizing the text information, the category information and the operation path;
the data cleaning unit is used for cleaning the data of the temporary storage database to obtain an atomic level text, wherein the atomic level text comprises a document title field, a category field and an operation path field;
and the atlas construction unit is used for constructing a knowledge atlas by using the atomic-level text, and the knowledge atlas takes the category field as an index.
In a third aspect, an embodiment of the present application provides an apparatus, where the apparatus includes a memory and a processor, where the memory is configured to store instructions or codes, and the processor is configured to execute the instructions or codes, so that the apparatus performs the method for recommending an operation path based on a knowledge-graph according to any one of the first aspect.
In a fourth aspect, an embodiment of the present application provides a computer storage medium, where code is stored, and when the code is executed, an apparatus that executes the code implements the operation path recommendation method based on a knowledge graph in any one of the foregoing first aspects.
The embodiment of the application provides an operation path recommending method based on a knowledge graph and a related product, wherein when the method is executed, identification information, category information and an operation path in a target product are acquired, the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence; constructing a knowledge graph by utilizing the identification information, the category information and the operation path; based on the knowledge graph, constructing a deep learning model, wherein the deep learning model is used for learning the relevance of the content in the knowledge graph; and inputting the user input corpus into the deep learning model, and outputting a recommended operation path.
In this way, the corresponding relation between the primary operation path and the operation related description is constructed through the knowledge graph, and then the relevance between the operation path and other operation related descriptions is learned through the deep learning model, so that when the corpus input by the user is obtained, the operation path corresponding to the operation path is quickly matched, the user does not need to manually find the required operation path in a large number of operation paths, and the acquisition efficiency of the operation path and the use experience of the user are improved.
Drawings
In order to more clearly illustrate this embodiment or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for recommending an operation path based on a knowledge graph according to an embodiment of the present application;
FIG. 2 is a flowchart of another method of recommending operation paths based on a knowledge graph according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an operation path recommendation method based on a knowledge graph according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of an operation path recommending apparatus based on a knowledge graph according to an embodiment of the present application.
Detailed Description
With the increasing number and updating of internet financial products, a large number of functions or programs are available for selection by users in each financial product. The user can hardly find the operation path of the function which the user wants to execute in a plurality of functions and programs, the use experience of the user can be influenced for a long time, and further the viscosity of the user is reduced, and the conversion rate of the product is reduced.
Meanwhile, when the internal seat personnel use the internal knowledge base to answer the customer problem, as the speed of updating the knowledge base is very high, the knowledge documents and the knowledge classification are increased, and the seat personnel can hardly find the answer content of the required problem, the seat personnel can possibly cause the customer waiting time to be longer, and the seat answering rate is reduced.
The method provided by the embodiment of the application is executed by the computer equipment and is used for improving the efficiency of finding the needed operation path for the user.
It should be noted that the operation path recommending method based on the knowledge graph and the related products provided by the application can be used in the artificial intelligence field or the financial field. The above is merely an example, and the application fields of the operation path recommending method based on the knowledge graph and the related products provided by the application are not limited.
It will be apparent that the described embodiments are only 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.
Referring to fig. 1, fig. 1 is a flowchart of a method for recommending an operation path based on a knowledge graph according to an embodiment of the present application, including:
step S101: and acquiring identification information, category information and an operation path in the target product.
The target product is an application program or a knowledge base. When the target product is an application program, the title name, the category and the front end URL can be obtained by grabbing a program name text library and are correspondingly used as identification information, category information and an operation path. When the target product is an application program, the document content, the document category and the URL address field information where the document is located can be obtained by capturing the document in the knowledge base, and the document category and the URL address field information are correspondingly used as identification information, category information and operation paths.
It should be noted that there may be one or more operation paths in the same target product, where the identification information, the category information and the operation paths are in one-to-one correspondence, that is, each operation path has identification information and category information corresponding to the operation paths in a group.
Step S102: and constructing a knowledge graph by using the identification information, the category information and the operation path.
Knowledge Graph (knowledgegraph), a large-scale semantic network, contains entities, concepts and various semantic relationships between them. Knowledge graph technology provides a better ability to organize, manage and understand the vast information of the internet, and the information of the internet is expressed in a form closer to the human cognitive world. A knowledge base with semantic processing capability and open interconnection capability is established, and application value can be generated in intelligent information services such as intelligent search, intelligent question-answering, personalized recommendation and the like.
By using the identification information, the category information and the operation path, a knowledge graph can be constructed, and the identification information, the category information and the operation path are in one-to-one correspondence, so that the knowledge graph can be formed by using the identification information, the category information and the operation path as contacts.
Step S103: and constructing a deep learning model based on the knowledge graph.
The knowledge graph can only reflect the acquired group corresponding identification information, category information and operation path relation, but through analyzing the relation and characteristics of the identification information and the category information, the association between the operation path and other information can be obtained, and the operation path with strong directivity corresponding to a certain identification information is analyzed. Therefore, the relevance of the content in the knowledge-graph can be learned by constructing a deep learning model. Thus, the recommended operation path is more accurate and perfect.
Step S104: and inputting the user input corpus into the deep learning model, and outputting a recommended operation path.
Because the deep learning model already comprises all correlations between the operation path and other operation related descriptions, when the user input corpus is obtained, the identification information and the category information associated with the user input corpus can be correspondingly matched, so that the recommended operation path is obtained, the time and effort for confirming the operation path manually are reduced, and the use experience of the user is improved.
In summary, according to the embodiment, through the deep learning model, the deep learning of the knowledge graph is performed, the recommendation method is constructed, and after the user inputs the corpus, the optimal operation path and the recommended operation path can be fed back according to the input text data, so that the acquisition efficiency of the operation path and the use experience of the user are improved.
In the embodiment of the present application, there are a plurality of possible implementations of the steps described in fig. 1, and the following descriptions are provided separately. It should be noted that the implementations presented in the following description are only exemplary and not representative of all implementations of the embodiments of the present application.
Referring to fig. 2, the flowchart of another method of the operational path recommendation method based on a knowledge graph according to an embodiment of the present application includes:
step S201: and acquiring identification information, category information and an operation path in the target product.
As one possible implementation, the web crawler may be utilized to capture identification information, category information, and operation paths in the target product. The web crawler can conveniently acquire a large number of data sources, so that deeper and more effective data analysis can be performed to obtain more value.
Step S202: and generating a temporary storage database by using the text information, the category information and the operation path.
Since the text information, category information and data amount of the operation path may be large, direct processing is inconvenient, so that it may be stored in a temporary database for further processing.
Step S203: and performing data cleaning on the temporary storage database to obtain an atomic level text.
The acquired data is directly stored in the temporary storage database, so the data in the temporary storage database is coarse-grained text data. And data cleaning is carried out on the temporary storage database, data conforming to the construction knowledge graph can be output, and the atomic level content of the data comprises a document title field, a category field and an operation path field, and other data information related to the category is attached.
Step S204: and constructing a knowledge graph by utilizing the atomic level text.
The atomic level text can be utilized to construct the knowledge graph because the atomic level text meets the construction requirement of the knowledge graph. It should be noted that, the knowledge graph may use the category field as an index. The knowledge-graph may be a knowledge-graph reflecting that text (e.g., a field) is associated with the operation path.
Step S205: and constructing a deep learning model based on the knowledge graph.
As a possible implementation manner, the knowledge graph may be first subjected to deep learning of named entity recognition, so as to obtain the identification information, the category information and the entity type of the operation path respectively; disambiguating and integrating the entity types to obtain the identification information, the category information and the common category of the operation path; then, based on the identification information, the category information and the common category of the operation path, establishing a corresponding relation of the identification information, the category information and the operation path; and finally, generating a deep learning model by utilizing the corresponding relation among the identification information, the category information and the operation path.
Wherein named entity recognition (Named Entity Recognition, NER) is intended to recognize an entity in a string of text and to label the type it refers to, such as a person name, place name or product name, etc. Specifically, according to the Message Understanding Conference (MUC) conference specification, the named entity recognition task includes three sub-tasks: entity name: name of person, place, product, etc.; the time expression: date, time, duration, etc.; the numerical expression: percentages, metrics, money, cardinalities, etc. For example, new edition S bank mobile phone bank gives free month of transfer in new spring gift-in 2022, in this sentence, "S bank" is a name of organization, "transfer" is a name of product, "free" is a name of activity information, and named entity recognition task can help us automatically find these entities by modeling.
Based on the deep learning of named entity recognition, entity types with strong directivity, such as entity types, time types and keyword types, can be analyzed, namely different entity types can be generated corresponding to the identification information, the category information and the operation path. And classifying, disambiguating and integrating the text data in the database after deep learning to obtain the common category among the identification information, the category information and the common category of the operation path. Based on the identification information, the category information and the common category of the operation path, the corresponding relation of the identification information, the category information and the operation path is established, then the corresponding relation is utilized to generate a deep learning model, new path association information and program operation relation information between entities can be deduced, and a complete operation path recommendation knowledge graph is constructed.
It should be noted that, in order to obtain a more accurate operation path recommendation result, an input corpus input by a user at a website may be obtained, a user corpus database may be constructed, and an initial deep learning model may be trained based on the user corpus database. In this way, the accuracy of the recommended operation path output by the deep learning model can be continuously optimized.
Step S206: and acquiring an operation path continuously accessed by the user.
The operation path that the user continuously accesses refers to the next operation path that the user will frequently access after accessing a certain operation path. When one operation is performed in the product, the linkage path can be recommended according to the operation performed by other users after the operation, and the operation of the user can be analyzed and intelligently fed back.
Step S207: and training the deep learning model based on the operation path continuously accessed by the user to obtain an iterative learning model.
The iterative learning model is used for learning the relevance of the operation paths continuously accessed by the user, so that linkage of the operation paths can be formed, when the first operation is carried out, the next recommended operation path can be popped up without inputting keywords by the user, and the user experience is improved fully.
Step S208: and responding to the access of a user to a first operation path, inputting the first operation path into the iterative deep learning model, and outputting a linkage operation path.
The first operation path is any one of the operation paths acquired in step S201. Since the iterative deep learning model can already associate the operation path with the predicted operation path, the operation path required for the next operation of the user can be pushed in time.
Step S209: and inputting the user input corpus into the deep learning model, and outputting a recommended operation path.
Step S210: and responding to the recommended operation path as a recommended document, and extracting and outputting key language segments of the recommended document.
When the user is an internal seat person, a situation that the recommended operation path is a recommended document is often encountered. When the recommended operation path is a recommended document, in order to facilitate the user to quickly preview the long text, the important points and terms such as the document and the like can be automatically extracted and fed back through deep learning, so that the user is helped to quickly preview the long text, the follow-up operation is assisted to judge, and the working efficiency and the market competitiveness of staff are improved.
The steps can be simplified into a block diagram referring to fig. 3, firstly, capturing by a crawler program to obtain coarse-granularity text data, then, cleaning the data, facilitating knowledge graph construction, and then, constructing a deep learning model according to the knowledge graph to obtain a model capable of recommending a product operation path.
In summary, in this embodiment, a document atomic level database and a document knowledge graph are constructed by using named entity recognition and knowledge graph technology. Based on the knowledge graph, the next operation to be performed by the user is quickly recommended and predicted by searching for keywords in the document knowledge graph. When one operation is performed in the product, the linkage path URL can be recommended according to the operation performed by other users after the operation, and the operation of the user can be analyzed and intelligently fed back. Through deep learning, key points and terms such as documents are automatically extracted and fed back, a user is helped to preview long texts rapidly, follow-up operation is assisted to judge, working efficiency and market competitiveness of staff are improved, the user does not need to manually find a required operation path in a large number of operation paths, and the acquisition efficiency of the operation paths and the use experience of the user are improved.
The embodiment of the application provides some specific implementation modes of the operation path recommending method based on the knowledge graph, and based on the specific implementation modes, the application also provides a corresponding device. The apparatus provided by the embodiment of the present application will be described in terms of functional modularization.
Referring to the schematic structural diagram of the operation path recommending device based on the knowledge graph shown in fig. 4, the device includes an acquisition module 401, a graph construction module 402, a model construction module 403, and a path recommending module 404.
The obtaining module 401 is configured to obtain identification information, category information and an operation path in a target product, where the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence;
a graph construction module 402, configured to construct a knowledge graph using the identification information, the category information, and the operation path;
a model building module 403, configured to build a deep learning model based on the knowledge graph, where the deep learning model is used to learn relevance of content in the knowledge graph;
and the path recommending module 404 is used for inputting the corpus input by the user into the deep learning model and outputting a recommended operation path.
As a possible implementation manner, the map building module 402 includes:
the database construction unit is used for generating a temporary storage database by utilizing the text information, the category information and the operation path;
the data cleaning unit is used for cleaning the data of the temporary storage database to obtain an atomic level text, wherein the atomic level text comprises a document title field, a category field and an operation path field;
and the atlas construction unit is used for constructing a knowledge atlas by using the atomic-level text, and the knowledge atlas takes the category field as an index.
As a possible implementation manner, the model building module 403 includes:
the naming identification unit is used for performing deep learning of naming entity identification on the knowledge graph to respectively obtain the identification information, the category information and the entity type of the operation path;
the eliminating unit is used for eliminating and integrating the entity types to obtain the common categories of the identification information, the category information and the operation paths;
the relation establishing unit is used for establishing the corresponding relation between the identification information, the category information and the operation path based on the identification information, the category information and the common category of the operation path;
and the model construction unit is used for generating a deep learning model by utilizing the corresponding relation among the identification information, the category information and the operation path.
As a possible implementation manner, the device further comprises:
the behavior acquisition module is used for acquiring an operation path continuously accessed by a user;
the iteration module is used for training the deep learning model based on the operation paths continuously accessed by the user to obtain an iteration learning model, and the iteration learning model is used for learning the relevance of the operation paths continuously accessed by the user;
and the linkage module is used for responding to the access of the user to the first operation path, inputting the first operation path into the iterative deep learning model and outputting the linkage operation path.
As a possible implementation manner, the device further comprises:
and the key module is used for responding to the recommended operation path as a recommended document, extracting and outputting key speech segments of the recommended document.
As a possible implementation manner, the obtaining module 401 includes:
and the crawler unit is used for utilizing the web crawler to grasp the identification information, the category information and the operation path in the target product.
The embodiment of the application also provides corresponding equipment and a computer storage medium, which are used for realizing the scheme provided by the embodiment of the application.
The device comprises a memory and a processor, wherein the memory is used for storing instructions or codes, and the processor is used for executing the instructions or codes so that the device can execute the operation path recommending method based on the knowledge graph according to any embodiment of the application.
The computer storage medium stores codes, and when the codes are executed, equipment for executing the codes realizes the operation path recommending method based on the knowledge graph according to any embodiment of the application.
The "first" and "second" in the names of "first", "second" (where present) and the like in the embodiments of the present application are used for name identification only, and do not represent the first and second in sequence.
From the above description of embodiments, it will be apparent to those skilled in the art that all or part of the steps of the above described example methods may be implemented in software plus general hardware platforms. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a storage medium, such as a read-only memory (ROM)/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network communication device such as a router) to perform the method according to the embodiments or some parts of the embodiments of the present application.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments for relevant points. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present application without undue burden.
The foregoing description of the exemplary embodiments of the application is merely illustrative of the application and is not intended to limit the scope of the application.

Claims (10)

1. An operation path recommending method based on a knowledge graph is characterized by comprising the following steps:
acquiring identification information, category information and an operation path in a target product, wherein the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence;
constructing a knowledge graph by utilizing the identification information, the category information and the operation path;
based on the knowledge graph, constructing a deep learning model, wherein the deep learning model is used for learning the relevance of the content in the knowledge graph;
and inputting the user input corpus into the deep learning model, and outputting a recommended operation path.
2. The method of claim 1, wherein constructing a knowledge-graph using the text information, category information, and operation path comprises:
generating a temporary storage database by using the text information, the category information and the operation path;
performing data cleaning on the temporary storage database to obtain an atomic level text, wherein the atomic level text comprises a document title field, a category field and an operation path field;
and constructing a knowledge graph by using the atomic level text, wherein the knowledge graph takes the category field as an index.
3. The method of claim 1, wherein the constructing a deep learning model based on the knowledge-graph comprises:
performing deep learning of named entity recognition on the knowledge graph to respectively obtain the identification information, the category information and the entity type of the operation path;
disambiguating and integrating the entity types to obtain the identification information, the category information and the common category of the operation path;
based on the identification information, the category information and the common category of the operation path, establishing a corresponding relation of the identification information, the category information and the operation path;
and generating a deep learning model by using the corresponding relation among the identification information, the category information and the operation path.
4. The method according to claim 1, wherein the method further comprises:
acquiring an operation path continuously accessed by a user;
training the deep learning model based on the operation paths continuously accessed by the user to obtain an iterative learning model, wherein the iterative learning model is used for learning the relevance of the operation paths continuously accessed by the user;
and responding to the access of a user to a first operation path, inputting the first operation path into the iterative deep learning model, and outputting a linkage operation path.
5. The method of claim 1, wherein after said inputting a user input corpus into said deep learning model to output a recommended operational path, said method further comprises:
and responding to the recommended operation path as a recommended document, and extracting and outputting key language segments of the recommended document.
6. The method of claim 1, wherein the obtaining the identification information, the category information, and the operation path in the target product comprises:
and capturing identification information, category information and an operation path in the target product by utilizing the web crawler.
7. An operation path recommending device based on a knowledge graph, wherein the device comprises:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring identification information, category information and an operation path in a target product, the target product is an application program or a knowledge base, and the identification information, the category information and the operation path are in one-to-one correspondence;
the map construction module is used for constructing a knowledge map by utilizing the identification information, the category information and the operation path;
the model construction module is used for constructing a deep learning model based on the knowledge graph, and the deep learning model is used for learning the relevance of the content in the knowledge graph;
and the path recommending module is used for inputting the corpus input by the user into the deep learning model and outputting a recommended operation path.
8. The apparatus of claim 7, wherein the map construction module comprises:
the database construction unit is used for generating a temporary storage database by utilizing the text information, the category information and the operation path;
the data cleaning unit is used for cleaning the data of the temporary storage database to obtain an atomic level text, wherein the atomic level text comprises a document title field, a category field and an operation path field;
and the atlas construction unit is used for constructing a knowledge atlas by using the atomic-level text, and the knowledge atlas takes the category field as an index.
9. An apparatus comprising a memory for storing instructions or code and a processor for executing the instructions or code to cause the apparatus to perform the knowledge-graph based operational path recommendation method of any one of claims 1 to 6.
10. A computer storage medium having code stored therein, which when executed, causes a computer storage device executing the code to implement the knowledge-graph-based operation path recommendation method of any one of claims 1 to 6.
CN202310629604.1A 2023-05-30 2023-05-30 Knowledge graph-based operation path recommendation method and related products Pending CN116644172A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310629604.1A CN116644172A (en) 2023-05-30 2023-05-30 Knowledge graph-based operation path recommendation method and related products

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310629604.1A CN116644172A (en) 2023-05-30 2023-05-30 Knowledge graph-based operation path recommendation method and related products

Publications (1)

Publication Number Publication Date
CN116644172A true CN116644172A (en) 2023-08-25

Family

ID=87618443

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310629604.1A Pending CN116644172A (en) 2023-05-30 2023-05-30 Knowledge graph-based operation path recommendation method and related products

Country Status (1)

Country Link
CN (1) CN116644172A (en)

Similar Documents

Publication Publication Date Title
US20210019341A1 (en) Implementing a software action based on machine interpretation of a language input
US20170262783A1 (en) Team Formation
CN104216913A (en) Problem answering frame
Alexander et al. Natural language web interface for database (NLWIDB)
WO2013071305A2 (en) Systems and methods for manipulating data using natural language commands
US11651015B2 (en) Method and apparatus for presenting information
CN116244344B (en) Retrieval method and device based on user requirements and electronic equipment
KR20190059084A (en) Natural language question-answering system and learning method
CN111563141A (en) Natural language for structured query generation via restitution
JP2020135135A (en) Dialog content creation assisting method and system
CN111190920A (en) Data interactive query method and system based on natural language
CN112507089A (en) Intelligent question-answering engine based on knowledge graph and implementation method thereof
Zouaoui et al. A novel quranic search engine using an ontology-based semantic indexing
Bazaga et al. Translating synthetic natural language to database queries with a polyglot deep learning framework
US11620282B2 (en) Automated information retrieval system and semantic parsing
CN114647719A (en) Question-answering method and device based on knowledge graph
CN117271558A (en) Language query model construction method, query language acquisition method and related devices
CN117149804A (en) Data processing method, device, electronic equipment and storage medium
Iacob et al. MARAM: tool support for mobile app review management.
CN116644172A (en) Knowledge graph-based operation path recommendation method and related products
US20220334808A1 (en) Methods, systems, and computer readable media for creating and using minimum dictionary language (mdl) to access data in closed-domain data sets
CN111753540B (en) Method and system for collecting text data to perform Natural Language Processing (NLP)
Abrosimova et al. The ontology-based event mining tools for monitoring global processes
CN110457435A (en) A kind of patent novelty analysis system and its analysis method
English et al. Back-propagation training of a neural network for word spotting

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