CN113392212A - Service knowledge graph construction method and device, electronic equipment and storage medium - Google Patents

Abstract

The present application provides a service knowledge graph construction method, device, electronic device and storage medium, which relate to the technical field of artificial intelligence. The embodiment of the present application obtains each target service text information for a search service scenario; determines at least one derivative service text information associated with each target service text information according to the first text feature of each target service text information; The association relationship between the information and at least one derivative service text information associated with each other, and the corresponding service knowledge graph is constructed; because the service knowledge graph construction solution provided by the embodiment of this application is not limited to the acquired target service text information, according to The target service text information can obtain the associated derivative service text information, so that the constructed service knowledge graph contains more abundant information.

Description

A service knowledge graph construction method, device, electronic device and storage medium

technical field

The present application relates to the field of computer technology, and in particular, to a method, apparatus, electronic device and storage medium for constructing a service knowledge graph.

Background technique

Knowledge graph refers to a tree-like topology network with entities and concepts as nodes and relationships between nodes as edges. The knowledge graph makes knowledge acquisition more direct, thus realizing the convenience, intelligence and humanization of data search.

When searching for information, information related to the search keyword input by the user can be searched according to a pre-built knowledge graph. At present, the manual construction method is adopted when constructing the knowledge graph. A large number of keywords are obtained manually, and the obtained keywords are used as the nodes of the knowledge graph; and the relationship between the two nodes is manually set. The nodes are connected as the edges of the knowledge graph; the construction of the knowledge graph is completed after the nodes and edges are generated. However, the amount and type of information contained in the knowledge graph obtained by artificial construction have great limitations, and the constructed knowledge graph is too single.

SUMMARY OF THE INVENTION

The present application provides a method, apparatus, electronic device and storage medium for constructing a service knowledge graph, which are used to flexibly construct a service knowledge graph with rich information.

In a first aspect, the present application provides a method for constructing a service knowledge graph, including:

For search service scenarios, obtain text information of each target service;

Determine at least one derivative service text information associated with each target service text information according to the first text feature of each target service text information respectively;

A corresponding service knowledge graph is constructed according to the association relationship between the respective target service text information and at least one derivative service text information associated with each other.

In a second aspect, the present application provides a service knowledge graph construction device, including:

an obtaining unit, used for obtaining text information of each target service according to the search service scenario;

a determining unit, configured to determine at least one derivative service text information associated with each target service text information according to the first text feature of each target service text information;

The processing unit is configured to construct a corresponding service knowledge graph according to the association relationship between the respective target service text information and at least one derivative service text information associated with each other.

Optionally, the processing unit is specifically used for:

Perform the following operations on each of the target service text information:

According to one target service text information in the respective service text information and at least one associated derivative service text information, the triple information including the target service text information, the association type, and the derivative service text information is respectively generated; wherein, the three The number of tuple information is the same as the number of derivative service text information associated with the one target service text information, and the association type is the target service text information and the derivative service text contained in the triple information. The type of relationship between information.

Optionally, the determining unit is also used for:

Determine the service type to which each target service text information belongs respectively according to the second text feature of each target service text information; and determine each derivative service text information according to the obtained third text feature of each derivative service text information The service type to which the service text information belongs;

The processing unit is also used to:

After the corresponding service knowledge graph is constructed, in response to the instruction for displaying the service knowledge graph, the corresponding nodes are generated according to the target service text information and the derived service text information contained in the generated triple information, respectively, and based on The respective service types corresponding to the respective target service text information and the respective service types corresponding to the respective derived service text information are classified and displayed on the generated corresponding nodes; The included association type generates the edges between each node; according to the generated nodes and the edges between the nodes, the constructed service knowledge graph is displayed.

Optionally, the derivative service text information includes synonymous text information that is in a synonymous relationship with the target service text information;

The determining unit is also used for:

Before determining at least one derivative service text information associated with each target service text information, respectively determine the text content information corresponding to each target service text information; The information content of the information is fused respectively to obtain the first text feature of the corresponding target service text information; and, the corresponding text content information of each candidate text information in the preset candidate text information set is determined; The information contents of the corresponding text content information are respectively fused to obtain the third text feature of the corresponding candidate text information; wherein the text content information includes any one of edit distance, semantic distance, co-occurrence information, and attribute information. species or combination;

The determining unit is specifically used for:

The following operations are respectively performed for each target service text information: based on the trained text correlation model, according to the first text feature of one target service text information in the various target service text information, and the preset candidate text information The third text feature of each candidate text information in the set determines the similarity between the one target service text information and each candidate text information in the preset candidate text information set; At least one synonymous text information that is synonymous with the one target service text information is screened out from the preset candidate text information set.

Optionally, the derivative service text information includes upper text information of the target service text information;

The processing unit is specifically used for:

According to the first text feature of each target service text information and the preset matching rule, at least one upper-level text information associated with each target service text information is screened from the preset candidate text information set.

Optionally, the obtaining unit is specifically used for:

Obtain each service text information historically input by multiple objects in the search client, and use the acquired service text information as the target service text information for the search service scenario; or

Each service text information is acquired from at least one database as the target service text information for the search service scenario.

Optionally, the processing unit is also used for:

In response to the search instruction triggered by the target object, obtain the search service text information contained in the search instruction; according to the constructed service knowledge graph, determine the search service text information from the service knowledge graph, and Searching for at least one piece of service text information associated with the service text information; using the search service text information and at least one service text information associated with the search service text information as the service text information to be recommended; The page display information corresponding to the text information of the service to be recommended is recommended to the target object.

In a third aspect, an embodiment of the present application provides an electronic device, including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the service knowledge graph construction method provided by the present application.

In a fourth aspect, an embodiment of the present application provides a computer-readable medium storing computer-executable instructions, where the computer-executable instructions are used to execute the service knowledge graph construction method provided by the present application.

Beneficial effects of this application:

In the solution for building a service knowledge graph provided by the embodiments of the present application, when building a service knowledge graph, first obtain a plurality of target service text information, and then determine at least one derivative service associated with each target service text information according to the text features of the target service text information Therefore, at least one associated derivative service text can be obtained according to a target service text information, so that the service text information can be enriched. In addition, after obtaining at least one derivative service text information associated with each target service text information, a corresponding service knowledge graph is constructed according to the association relationship between each target service text information and the at least one derivative service text information associated with each target service text information. Therefore, the service knowledge graph construction solution provided by the embodiments of the present application is not limited to the acquired target service text information, and related derivative service text information can be obtained according to the target service text information, so that the constructed service knowledge graph contains more for rich information.

Description of drawings

The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the description, serve to explain the principles of the present application, and do not constitute an improper limitation of the present application.

1 is a schematic diagram of an application scenario of a method for constructing a service knowledge graph according to an embodiment of the present application;

2 is a flowchart of a method for constructing a service knowledge graph according to an embodiment of the present application;

FIG. 3 is a schematic diagram of determining text information of a derivative service provided by an embodiment of the present application;

FIG. 4 is another schematic diagram of determining text information of derivative services provided by an embodiment of the present application;

FIG. 5 is a schematic diagram of a service knowledge graph provided by an embodiment of the present application;

FIG. 6 is a schematic diagram of an application scenario of a service knowledge graph provided by an embodiment of the present application;

FIG. 7 is a schematic flowchart of a search method based on a service knowledge graph provided by an embodiment of the present application;

FIG. 8 is a schematic structural diagram of an apparatus for constructing a service knowledge graph according to an embodiment of the present application;

FIG. 9 is a schematic structural diagram of an electronic device provided by an embodiment of the present application;

FIG. 10 is a schematic structural diagram of a computing device according to an embodiment of the present application.

Detailed ways

In order to make those skilled in the art better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first", "second", etc. in the description and claims of the present application and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. It is to be understood that data so used may be interchanged under appropriate circumstances so that the embodiments of the application described herein can be practiced in sequences other than those illustrated or described herein. The implementations described in the illustrative examples below are not intended to represent all implementations consistent with this application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as recited in the appended claims.

Hereinafter, some terms in the embodiments of the present application will be explained, so as to facilitate the understanding of those skilled in the art.

1. Vertical search: Vertical search is a professional search engine for a certain industry. It is a subdivision and extension of the search engine. It integrates a certain type of specialized information in the library, and extracts the required data for processing by directional sub-fields. It is then returned to the user in some form. For example, official account search, small program search, etc.

2. Service search: a user can search for a service through a search engine, the user enters a search keyword (query), and the web page or client can directly display the service satisfying the user's search keyword to the user. For example, when a user searches for a babysitter, the service search can directly provide a menu of services for hiring babysitters. Service search is a type of vertical search.

3. Knowledge graph: It combines the theories and methods of applied mathematics, graphics, information visualization technology, information science and other disciplines with methods such as metrology citation analysis and co-occurrence analysis, and uses a visual graph to visually display the core of the discipline The structure, development history, frontier fields and overall knowledge structure achieve modern theory for the purpose of multidisciplinary integration. The complex knowledge field is displayed through data mining, information processing, knowledge measurement and graphic drawing, revealing the dynamic development law of the knowledge field, and providing practical and valuable reference for subject research. Through the effective processing, processing and integration of the data of the intricate documents, it is transformed into a simple and clear "entity-relation-entity" triplet, and a topology map that aggregates a large amount of knowledge.

4. Natural Language Processing (NLP): It is an important direction in the field of computer science and artificial intelligence. It studies various theories and methods that can realize effective communication between humans and computers using natural language. Natural language processing is a science that integrates linguistics, computer science, and mathematics.

5. Semantics: The information contained in language is semantics. Symbols are the carrier of language. Symbols themselves have no meaning. Only symbols that have been given meaning can be used. At this time, language is transformed into information, and the meaning of language is Semantic.

6. Terminal: Also known as User Equipment (UE), Mobile Station (MS), Mobile Terminal (MT), etc., it is a device that provides voice and/or data connectivity to users , for example, handheld devices with wireless connectivity, in-vehicle devices, etc. At present, some examples of terminals are: a mobile phone (mobile phone), a tablet computer, a notebook computer, a palmtop computer, and a mobile Internet device (Mobile Internet Device, MID).

7. Client: It can refer to either a software application (Application, APP) or a terminal device. It has a visual display interface and can interact with users; it corresponds to the server and provides local services for customers. For software applications, except for some applications that only run locally, they are generally installed on common client terminals and need to run in cooperation with the server. After the development of the Internet, the more commonly used applications include e-mail clients for sending and receiving e-mails, and instant messaging clients. For this type of application, there needs to be a corresponding server and service program in the network to provide corresponding services, such as database service, configuration parameter service, etc., so that a specific communication connection needs to be established between the client terminal and the server to ensure the application normal operation of the program.

8. Server: It can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or it can provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud Cloud servers for basic cloud computing services such as communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms.

The design ideas of the embodiments of the present application are briefly introduced below:

Artificial Intelligence (AI) is a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machines have the functions of perception, reasoning and decision-making.

Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, including both hardware-level technology and software-level technology. The basic technologies of artificial intelligence generally include technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation/interaction systems, and mechatronics. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning/deep learning.

Machine Learning (ML) is the study of how computers simulate or implement human learning behaviors to acquire new knowledge or skills, and to reorganize existing knowledge structures to continuously improve their performance. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, teaching learning and other technologies. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications are in all fields of artificial intelligence.

With the continuous popularity of search engines and various clients, users can use the search function in more and more scenarios. Users enter search keywords in the search engine or client, and the search engine or client can be based on pre-built knowledge. Graph search displays information related to the search keywords entered by the user to the user. At present, the manual construction method is adopted when constructing the knowledge graph. A large number of keywords are obtained manually, and the obtained keywords are used as the nodes of the knowledge graph; and the relationship between the two nodes is manually set. The nodes are connected as the edges of the knowledge graph; the construction of the knowledge graph is completed after the nodes and edges are generated. However, the amount and type of information contained in the knowledge graph obtained by artificial construction have great limitations. Moreover, the currently constructed knowledge graphs are all aimed at the item or commodity search scenario. Based on the constructed knowledge graph, other items or commodities similar to the user’s searched item or commodity are recommended to the user; for example, the search keyword entered by the user is “” Air Conditioning”, it will recommend air conditioners of various brands and functions to users. Currently, there is no service knowledge graph built for service scenarios.

Therefore, an embodiment of the present application provides a method for constructing a service knowledge graph, which acquires text information of each target service according to a search service scenario; At least one derivative service text information; according to the association relationship between each target service text information and at least one derivative service text information associated with each, a corresponding service knowledge graph is constructed. In the solution for building a service knowledge graph provided by the embodiments of the present application, when building a service knowledge graph, first obtain a plurality of target service text information, and then determine at least one derivative service associated with each target service text information according to the text features of the target service text information Therefore, at least one associated derivative service text can be obtained according to a target service text information, so that the service text information can be enriched. In addition, after obtaining at least one derivative service text information associated with each target service text information, a corresponding service knowledge graph is constructed according to the association relationship between each target service text information and the at least one derivative service text information associated with each target service text information. Therefore, the service knowledge graph construction solution provided by the embodiments of the present application is not limited to the acquired target service text information, and related derivative service text information can be obtained according to the target service text information, so that the constructed service knowledge graph contains more for rich information.

After introducing the design ideas of the embodiments of the present application, the following briefly introduces some application scenarios to which the technical solutions of the embodiments of the present application can be applied. It should be noted that the application scenarios introduced below are only used to illustrate the embodiments of the present application and not limited. In the specific implementation process, the technical solutions provided by the embodiments of the present application may be flexibly applied according to actual needs.

As shown in FIG. 1, it is a schematic diagram of an exemplary application scenario of an embodiment of the present application, including a user 10, a terminal 11, and a server 12;

An optional application scenario is that the user 10 triggers the construction instruction of the service knowledge graph through the display interface of the terminal 11, and the terminal 11 sends the construction instruction triggered by the user 10 to the server 12; the server 12 obtains the construction instruction after receiving the construction instruction. For each target service text information of the search service scenario; according to the first text feature of each target service text information, determine at least one derivative service text information associated with each target service text information; The association relationship between at least one derivative service text information is used to construct a corresponding service knowledge graph; after the server 12 constructs and obtains the service knowledge graph, it stores the constructed service knowledge graph.

The server in this embodiment of the present application may be an independent server, or may be a server cluster composed of multiple servers. The terminal 11 and the server 12 are connected through a network for communication, and the network may be a local area network, a wide area network, or the like.

The following describes the service knowledge graph construction method provided by the exemplary embodiment of the present application with reference to FIG. 2 to FIG. 5 in conjunction with the application scenarios described above. It should be noted that the above application scenarios are only shown for the convenience of understanding the spirit and principles of the present application, and the embodiments of the present application are not limited in this respect. Rather, the embodiments of the present application can be applied to any scenario where applicable.

As shown in FIG. 2 , a schematic flowchart of a method for constructing a service knowledge graph provided in an embodiment of the present application, the method may include the following steps:

Step S201, for the search service scenario, obtain text information of each target service;

Step S202: Determine at least one derivative service text information associated with each target service text information according to the first text feature of each target service text information;

Step S203 , constructing a corresponding service knowledge graph according to the association relationship between each target service text information and at least one associated derivative service text information.

The target service text information in this embodiment of the present application may be keywords with service search intent;

The acquisition methods of the target service text information in this embodiment of the present application may include but are not limited to the following methods:

Mode 1: Acquire each service text information historically input by multiple objects in the search client, and use the acquired service text information as the target service text information for the search service scenario;

In this way, a large number of service text information historically input by users in the search client can be collected; since the service text information historically input by the user in the search client are generally keywords with service search intentions, it is possible to use The service text information historically input by a large number of users in the search client is used as the target service text information for the search service scenario.

For example, the service text information historically input by the user in the search client may be text information such as "repair air conditioner", "repair washing machine", "dry cleaning", "find a nanny", "find a job" and so on.

Mode 2: Acquire each service text information from at least one database as the target service text information for the search service scenario;

In this way, when constructing a service knowledge graph in this embodiment of the present application, service text information may be obtained from one or more databases, and the service text information obtained from one or more databases may be texts with service search intentions information.

When constructing a service knowledge graph in the embodiment of the present application, it is necessary to classify the nodes in the service knowledge graph according to the category system to which the nodes belong; after acquiring the text information of each target service, determine the service type to which the text information of each target service belongs. ;

It should be noted that the service type to which the target service text information belongs is the category system to which the target service text information belongs in the service knowledge graph.

In some embodiments, the service type to which each target service text information belongs is determined according to the following methods:

Based on the trained text classification model, determine the service type to which each target service text information belongs;

In implementation, each target service text information may be input into the trained text classification model respectively, and the output of the trained text classification model may be used as the service type to which each target service text information belongs.

Among them, the trained text classification model includes feature extraction network and classification network;

The feature extraction network of the trained text classification model randomly initializes the target service text information by embedding;

The classification network of the trained text classification model classifies the randomly initialized embedding vectors to obtain the service type to which each target service text information belongs.

The following describes in detail how to obtain the service type to which each target service text information belongs based on the trained text classification model:

Based on the feature extraction network of the trained text classification model, the following operations are performed for each target service text information:

Extract the word vector, word vector and position vector representing the relative positional relationship between characters in a target service text information from each target service text information; the extracted word vector, word vector and position vector Perform fusion processing to obtain a second text feature of the target service text information.

After obtaining the second text feature of the target service text information based on the feature extraction network of the trained text classification model, determine the service type to which each target service text information belongs based on the classification network of the trained text classification model;

Based on the classification network of the trained text classification model, the following operations are performed for each target service text information:

Perform averaging processing on the second text feature of a target service text information in each target service text information to obtain a classification result for representing the probability that a target service text information belongs to each preset service type; according to the classification result , to determine the service type to which a target service text message belongs.

It should be noted that, the text classification model in the embodiment of the present invention may be a Fasttext model.

In this embodiment of the present application, after acquiring each target service text information, the attribute information of each target service text information can also be determined;

When determining the attribute information of each target service text information, the embodiment of the present application acquires the attribute feature corresponding to each target service text information;

The attribute features corresponding to the target service text information include but are not limited to part-of-speech features, embedding features, and word frequency features;

For part-of-speech features, stammering word segmentation can be used to obtain the part-of-speech features of each target service text information; for example, the part-of-speech features can be nouns, verbs, gerunds, etc.;

For the embedding feature, the trained word2vec model can be used to obtain the embedding feature of the text information of each target service;

For word frequency features, the trained TF IDF model can be used to obtain word frequency features of each target service text information.

After the attribute features corresponding to the target service text information are obtained, the attribute features corresponding to each target service text information are input into the trained Xgboost classification model to obtain the attribute information of each target service text information.

In this embodiment of the present application, after acquiring each target service text information, according to the first text feature of each target service text information, at least one derivative service text information associated with each target service text information is determined;

It should be noted that, in this embodiment of the present application, each target service text information may determine at least one derivative service text information.

Wherein, the derivative service text information includes synonymous text information that is in a synonymous relationship with the target service text information, and superordinate text information of the target service text information;

The following describes the determination methods of these two types of derivative service text information:

1. The derivative service text information and the target service text information are synonymous text information with a synonymous relationship;

For example, when the target service text information is an entity word, the derived service text information is a synonym or extension of the entity word.

When determining the synonymous text information of the target service text information, the synonymous text information of each target service text information can be determined based on the trained text correlation model;

Based on the trained text correlation model, extract the first text feature of each target service text information;

In implementation, for each target service text information, the first text feature of each target service text information can be obtained according to the following methods:

The text content information corresponding to each target service text information is determined respectively; the information content of the text content information corresponding to each service text information is respectively fused to obtain the first text feature of the corresponding target service text information.

In addition, in the embodiment of the present application, when the synonymous text information of each target service text information is determined based on the trained text correlation model, it is based on the first text feature of each target service text information and the preset candidate text information set in the set of text information. For the third text feature corresponding to each candidate text information, at least one synonymous text information that is synonymous with the target service text information is selected from the preset candidate text information set;

In implementation, for each candidate text information in the preset candidate text information set, the third text feature of each candidate text information can be obtained in the following manner:

Determine the corresponding text content information of each candidate text information in the preset candidate text information set; respectively fuse the information content of the text content information corresponding to each candidate text information to obtain the third text feature of the corresponding candidate text information.

It should be noted that the text content information includes any one or a combination of edit distance, semantic distance, co-occurrence information, and attribute information;

Among them, the edit distance is a quantitative measure of the degree of difference between two character strings in text information, and the measurement method is to see how many times of processing is required to turn one character string into another character string. Edit distance can be used in NLP; semantic distance is the distance in semantic space; co-occurrence information is the phenomenon that the information described by text information co-occurs.

The way of determining the attribute information can be referred to the above description; for example, when the target service text information includes service words, the attribute information includes service entity words, service behavior words, service status words, service compound words, and the like.

After obtaining the first text feature of each target service text information and the third text feature of each candidate text information in the preset candidate text information set, based on the trained text correlation model, obtain the respective associations of each target service text information at least one derivative service text message of ;

For example, the text correlation model in this embodiment of the present application may be a linear LR model.

During implementation, perform the following operations for each target service text information:

Based on the trained text correlation model, a target is determined according to the first text feature of one target service text information in each target service text information, and the third text feature of each candidate text information in the preset candidate text information set The similarity between the service text information and each candidate text information in the preset candidate text information set; according to each determined similarity, screen out from the preset candidate text information set as synonymous with a target service text information At least one synonymous textual information of the association relationship.

It should be noted that, when the synonymous text information associated with the target text information is determined, synonymous text information of different attributes can be obtained according to the attribute information of the target text information and the attribute information of the candidate text information.

For example, when the target text information is "air conditioner", the corresponding synonym text information includes synonyms and extended words; wherein, the obtained synonym corresponding to "air conditioner" may be "refrigeration air conditioner", and the obtained synonym corresponding to "air conditioner" The expanded words can be automobile air conditioners, central air conditioners, air conditioners, air conditioners, air conditioners, air conditioners, air conditioners, inverter air conditioners, etc.

2. The derivative service text information and the target service text information are the upper text information of the target service text information;

For example, when the target service text information is an entity word, the derived service text information is a hypernym of the entity word.

In some embodiments, at least one upper-level text information associated with each target service text information is obtained according to the following methods:

According to the first text feature of each target service text information and the preset matching rule, at least one upper-level text information associated with each target service text information is screened from the preset candidate text information set.

It should be noted that, in this embodiment of the present application, when the associated upper text information is obtained according to the target service text information, a set of upper text information associated with the target service text information may be obtained based on manual pattern mining. For example, if pattern mining determines that A belongs to B, then B is the upper text information of A.

Through the method described above, for each target service text information, the service type to which the target service text information belongs, the attribute information of the target service text information, the synonymous text information associated with the target service text information, and the upper level of the target service text information can be obtained. text information.

For example, as shown in FIG. 3 , the target service text information in this embodiment of the present application includes air conditioners: determine the service type to which "air conditioner" belongs is housekeeping; determine the attribute information of "air conditioner" as a service entity word; determine the extension of "air conditioner" The words include automobile air conditioner, central air conditioner, air conditioner equipment, air conditioner, air conditioner vehicle, air conditioner network, air, inverter air conditioner; the synonym of "air conditioner" is determined as refrigeration air conditioner; the hypernym of "air conditioner" is determined to include household appliances, household appliances, household appliances , electrical appliances, consumer electronics.

As shown in FIG. 4 , the text information of the target service in this embodiment of the present application includes express: determine that the service type to which “express” belongs is housekeeping or transportation; The extended word of "fast" includes fast and speed; under the traffic type, the attribute information of "fast" is determined as the service status word, and the extended word of "fast" includes express road, speed and high speed.

In this embodiment of the present application, after determining at least one derivative service text information associated with each target service text information, a corresponding service knowledge graph is constructed according to the association relationship between each target service text information and at least one derivative service text information associated with each target service text information. ;

In some embodiments, the service knowledge graph can be constructed in the following manner:

According to one target service text information in each service text information and the associated at least one derivative service text information, the triple information including the target service text information, the association type, and the derived service text information is respectively generated; wherein, the triple information of the triple information The number is the same as the number of derivative service text information associated with a target service text information, and the association type is the type of the association relationship between the target service text information contained in the triplet information and the derivative service text information.

It should be noted that the service knowledge graph constructed in the embodiment of the present application may be stored in the form of triple information, and each triple information is represented in the form of <entity-relation-entity>; the entity may be the target Service text information or derivative service text information, the relationship is an association relationship between two entities;

Wherein, the association relationship between the two entities in this embodiment of the present application includes, but is not limited to: belonging to, acting on, behavior is, synonymous, located, brand is, main service is, service area, service crowd, and occurrence.

For example, if the target service text information is "washing machine repair", it is determined that the upper text information includes home appliance repair, and the synonymous text information is determined to include washing machines and repairs; the target service text information is "toilet repair", and it is determined that the upper text information includes house repair, kitchen and bathroom Maintenance, determine that the synonymous text information includes toilet and repair; the target service text information is "kitchen and bathroom maintenance", determine that the upper text information includes house maintenance, and determine that the synonymous text information includes kitchen and bathroom utensils, maintenance, and repair;

Then the generated triple information includes:

<Washing Machine Repair-belongs-Home Appliance Repair>, <Washing Machine Repair-Synonym-Washing Machine>, <Washing Machine Repair-Synonym-Maintenance>, <Toilet Repair-Belong-House Repair>, <Toilet Repair-Belong-Kitchen Maintenance> , <Toilet Repair-Synonym-Toilet>, <Toilet Repair-Synonym-Repair>, <Kitchen Maintenance-Belongs-House Maintenance>, <Kitchen Maintenance-Synonym-Kitchen Appliance>, <Kitchen Maintenance- Synonym-Maintenance>, <Kitchen Maintenance-Synonym-Repair>.

In addition, the embodiment of the present application also needs to determine the service type of each target service text information and the derived service text information, wherein, for the determination method of the service type to which the target service text information belongs, reference may be made to the above description, and the service type to which the derived service text information belongs. The manner of determining the type is similar to the manner of determining the service type to which the target service type belongs, and will not be described in detail here.

When displaying the constructed service knowledge graph, the service knowledge graph can be displayed in the form of a tree topology; an optional implementation is that when displaying the constructed service knowledge graph, it is classified and displayed according to different service types;

In the implementation, the service type to which each target service text information belongs is determined respectively; and the service type to which each derivative service text information belongs is determined respectively;

In response to the instruction to display the service knowledge graph, the corresponding nodes are generated according to the target service text information and the derived service text information contained in the generated triple information, and the corresponding service type and each Derive the corresponding service types of the service text information, and classify and display the corresponding generated nodes; and generate the edges between the nodes according to the association types contained in the generated triple information; The edge between them displays the constructed service knowledge graph.

For example, the service category system included under the housekeeping service type includes home appliance repair and house repair; among them, the triple information included under the home appliance repair service category system is as follows:

<Washing Machine Maintenance-Synonym-Washing Machine>, <Washing Machine Maintenance-Host-Maintenance>, <Washing Machine-Host-Home Appliances>, <Home Appliances-Synonym-Household Goods>;

The triple information included in the housing maintenance service category system is as follows:

<Toilet Repair-Synonym-Toilet>, <Toilet Repair-Superior-Kitchen Maintenance>, <Toilet Repair-Superior-Repair>, <Toilet-Superior-Kitchen and Bathroom Appliances>, <Kitchen and Bathroom Maintenance-Superior-Kitchen and Bathroom Appliances >, <kitchen and bathroom maintenance-synonym-repair>, <kitchen and bathroom maintenance-synonym-repair>, <kitchen and bathroom appliances-superior-household goods>, <maintenance-synonym-repair>, <maintenance-synonym- cleaning>, <repair-synonym-cleaning>.

The service category system included under the government service type includes traffic management service and household registration service; among them, the triple information included under the traffic management service category system is as follows:

<traffic violation-synonym-traffic violation inquiry>, <traffic violation inquiry-superior-traffic violation>, <traffic violation inquiry-superior-inquiry>, <traffic violation-superior-traffic>, <traffic violation-synonym-violation >;

The triple information included in the household registration service category system is as follows:

<ID card-synonym-ID card loss report>, <ID card-synonym-inquiry>, <ID card loss report-superior-ID card>, <ID card loss-superior-report loss>.

Based on the above triplet information, the generated service knowledge graph is shown in Figure 5.

After the service knowledge graph is constructed in this embodiment of the present application, the constructed service knowledge graph can be applied to a search scenario; an optional schematic diagram of an application scenario of the service knowledge graph is shown in FIG. 6 , including a user 60 , a terminal 61 and a server 62 ; Wherein, a client terminal is installed on the terminal 61;

The user 60 triggers the search instruction through the client installed on the terminal 61, and the client obtains the search service text information contained in the search instruction; and sends the search service text information to the server 62, and the server 62, according to the constructed service knowledge graph, obtains the information from the service knowledge. The search service text information and at least one service text information related to the search service text information are determined in the map; the search service text information and the at least one service text information related to the search service text information are used as to-be-recommended Service text information; the determined page display information corresponding to the text information of the service to be recommended is returned to the client, and the client displays the page display information corresponding to the text information of the service to be recommended to the target object in the display interface.

As shown in FIG. 7 , a schematic flowchart of a search method based on a service knowledge graph provided by an embodiment of the present application, the method may include the following steps:

Step S701, the client obtains the search service text information contained in the search instruction in response to the search instruction triggered by the target object;

Step S702, the client sends the search service text information to the server;

Step S703, the server determines the search service text information and at least one service text information associated with the search service text information from the service knowledge map according to the constructed service knowledge map;

Step S704, the server uses the search service text information and at least one service text information associated with the search service text information as the service text information to be recommended;

Step S705, the server returns the determined page parameter corresponding to the text information of the service to be recommended to the client;

Step S706, the client recommends the page display information corresponding to the text information of the service to be recommended to the target object.

Based on the same inventive concept, an embodiment of the present application also provides a title text processing device. Since the principle of the device for solving problems is similar to the above title text processing method, the implementation of the device can refer to the implementation of the method, and the repetition is not repeated. Repeat.

As shown in FIG. 8 , a schematic structural diagram of an apparatus for constructing a service knowledge graph provided by an embodiment of the present application includes:

an obtaining unit 801, configured to obtain text information of each target service for a search service scenario;

a determining unit 802, configured to determine at least one derivative service text information associated with each target service text information according to the first text feature of each target service text information;

The processing unit 803 is configured to construct a corresponding service knowledge graph according to the association relationship between the respective target service text information and at least one derivative service text information associated with each other.

Optionally, the processing unit 803 is specifically configured to:

Perform the following operations on each of the target service text information:

According to one target service text information in the respective service text information and at least one associated derivative service text information, the triple information including the target service text information, the association type, and the derivative service text information is respectively generated; wherein, the three The number of tuple information is the same as the number of derivative service text information associated with the one target service text information, and the association type is the target service text information and the derivative service text contained in the triple information. The type of relationship between information.

Optionally, the determining unit 802 is further configured to:

Determine the service type to which each target service text information belongs respectively according to the second text feature of each target service text information; and determine each derivative service text information according to the obtained third text feature of each derivative service text information The service type to which the service text information belongs;

The processing unit 803 is also used for:

After the corresponding service knowledge graph is constructed, in response to the instruction for displaying the service knowledge graph, the corresponding nodes are generated according to the target service text information and the derived service text information contained in the generated triple information, respectively, and based on The respective service types corresponding to the respective target service text information and the respective service types corresponding to the respective derived service text information are classified and displayed on the generated corresponding nodes; The included association type generates the edges between each node; according to the generated nodes and the edges between the nodes, the constructed service knowledge graph is displayed.

Optionally, the derivative service text information includes synonymous text information that is in a synonymous relationship with the target service text information;

The determining unit 802 is further configured to:

Before determining at least one derivative service text information associated with each target service text information, respectively determine the text content information corresponding to each target service text information; The information content of the information is fused respectively to obtain the first text feature of the corresponding target service text information; and, the corresponding text content information of each candidate text information in the preset candidate text information set is determined; The information contents of the corresponding text content information are respectively fused to obtain the third text feature of the corresponding candidate text information; wherein the text content information includes any one of edit distance, semantic distance, co-occurrence information, and attribute information. species or combination;

The determining unit 802 is specifically configured to:

The following operations are respectively performed for each target service text information: based on the trained text correlation model, according to the first text feature of one target service text information in the various target service text information, and the preset candidate text information The third text feature of each candidate text information in the set determines the similarity between the one target service text information and each candidate text information in the preset candidate text information set; At least one synonymous text information that is synonymous with the one target service text information is screened out from the preset candidate text information set.

Optionally, the derivative service text information includes upper text information of the target service text information;

The determining unit 802 is specifically configured to:

According to the first text feature of each target service text information and the preset matching rule, at least one upper-level text information associated with each target service text information is screened from the preset candidate text information set.

Optionally, the obtaining unit 801 is specifically used for:

Obtain each service text information historically input by multiple objects in the search client, and use the acquired service text information as the target service text information for the search service scenario; or

Each service text information is acquired from at least one database as the target service text information for the search service scenario.

Optionally, the processing unit 803 is further configured to:

In response to the search instruction triggered by the target object, obtain the search service text information contained in the search instruction; according to the constructed service knowledge graph, determine the search service text information from the service knowledge graph, and Searching for at least one piece of service text information associated with the service text information; using the search service text information and at least one service text information associated with the search service text information as the service text information to be recommended; The page display information corresponding to the text information of the service to be recommended is recommended to the target object.

For the convenience of description, the above parts are divided into modules (or units) according to their functions and described respectively. Of course, the functions of each module (or unit) may be implemented in one or more software or hardware when implementing the present application.

As will be appreciated by one skilled in the art, each aspect of the present application may be implemented as a system, method or program product. Therefore, each aspect of the present application can be embodied in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which may be collectively referred to herein as is a "circuit", "module" or "system".

In some possible implementations, the embodiments of the present application further provide an electronic device. Referring to FIG. 9 , the electronic device 900 may at least include at least one processor 901 and at least one memory 902 . Wherein, the memory 902 stores program codes, and when the program codes are executed by the processor 901, the processor 901 is made to execute the steps in the service knowledge graph construction method according to various exemplary embodiments of the present application described above in this specification, for example, The processor 901 may perform the steps shown in FIG. 2 or FIG. 7 .

In some possible implementation manners, the embodiments of the present application further provide a computing device, which may at least include at least one processing unit and at least one storage unit. The storage unit stores program codes, and when the program codes are executed by the processing unit, the processing unit executes the steps in the service knowledge graph construction method according to various exemplary embodiments of the present application described above in this specification, for example, a processor 901 may perform steps as shown in FIG. 2 or FIG. 7 .

A computing device 1000 according to such an embodiment of the present application is described below with reference to FIG. 10 . The computing device 1000 in FIG. 10 is only an example, and should not impose any limitations on the functions and scope of use of the embodiments of the present application.

As shown in FIG. 10, computing device 1000 takes the form of a general-purpose computing device. Components of the computing device 1000 may include, but are not limited to, the above-mentioned at least one processing unit 1001 , the above-mentioned at least one storage unit 1002 , and a bus 1003 connecting different system components (including the storage unit 1002 and the processing unit 1001 ).

Bus 1003 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a processor, or a local bus using any of a variety of bus structures.

The storage unit 1002 may include readable media in the form of volatile memory, such as random access memory (RAM) 1021 or cache memory 1022 , and may further include read only memory (ROM) 1023 .

The storage unit 1002 may also include a program/utility 1025 having a set (at least one) of program modules 1024 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, An implementation of a network environment may be included in each or some combination of these examples.

The computing apparatus 1000 may also communicate with one or more external devices 1004 (eg, keyboards, pointing devices, etc.), and may also communicate with one or more devices that enable a user to interact with the computing apparatus 1000, or communicate with the computing apparatus 1000. Any device (eg, router, modem, etc.) that communicates with one or more other computing devices. Such communication may take place through input/output (I/O) interface 1005 . Also, the computing device 1000 may communicate with one or more networks (eg, a local area network (LAN), a wide area network (WAN), or a public network such as the Internet) through a network adapter 1006 . As shown, network adapter 1006 communicates with other modules for computing device 1000 via bus 1003 . It should be understood that, although not shown, other hardware or software modules may be used in conjunction with computing device 1000, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data Backup storage systems, etc.

In some possible implementations, each aspect of the service knowledge graph construction method provided by the present application can also be implemented in the form of a program product, which includes program code, and when the program product runs on a computer device, the program code uses To make the computer device execute the steps in the service knowledge graph construction method according to various exemplary embodiments of the present application described above in this specification, for example, the computer device may execute the steps as shown in FIG. 2 or FIG. 7 .

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples (non-exhaustive list) of readable storage media include: electrical connections with one or more wires, portable disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

While the preferred embodiments of the present application have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be construed to include the preferred embodiment and all changes and modifications that fall within the scope of this application.

Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims (10)

1. A service knowledge graph construction method is characterized by comprising the following steps:

acquiring text information of each target service aiming at a search service scene;

determining at least one derivative service text message associated with each target service text message according to the first text feature of each target service text message;

and constructing a corresponding service knowledge graph according to the association relationship between each target service text message and each associated at least one derivative service text message.

2. The method of claim 1, wherein constructing a corresponding service knowledge graph according to the association relationship between each target service text message and the respectively associated at least one derived service text message comprises:

respectively executing the following operations aiming at each target service text message:

generating triple information containing target service text information, association type and derived service text information respectively according to one target service text information and associated at least one derived service text information in each service text information; the number of the triple information is the same as the number of the derived service text information associated with the target service text information, and the association type is the type of the association relationship between the target service text information and the derived service text information contained in the triple information.

3. The method of claim 2, wherein the method further comprises:

determining the service type of each target service text message according to the second text feature of each target service text message; determining the service type of each derived service text message according to the obtained third text feature of each derived service text message;

after the building of the corresponding service knowledge graph, the method further comprises:

responding to an instruction for displaying the service knowledge graph, generating corresponding nodes according to target service text information and derived service text information contained in each generated triple information, and performing classified display on the generated corresponding nodes based on service types corresponding to each target service text information and each derived service text information; and

generating edges among the nodes according to the generated association types contained in the triple information;

and displaying the constructed service knowledge graph according to the generated nodes and the edges among the nodes.

4. The method of claim 1, wherein the derived service text information comprises synonymous text information in a synonymous association with the target service text information;

before the determining at least one derivative service text message associated with each target service text message, the method further includes:

respectively determining text content information corresponding to each target service text information; respectively fusing the information content of the text content information corresponding to each service text information to obtain a first text characteristic of the corresponding target service text information; and

determining text content information corresponding to each candidate text information in a preset candidate text information set; respectively fusing the information content of the text content information corresponding to each candidate text information to obtain a third text characteristic of the corresponding candidate text information; the text content information comprises any one or combination of an editing distance, a semantic distance, co-occurrence information and attribute information;

the determining at least one derivative service text message associated with each target service text message according to the first text feature of each target service text message respectively includes:

respectively executing the following operations aiming at each target service text message: based on a trained text correlation model, determining similarity between one target service text message and each candidate text message in the preset candidate text message set according to a first text feature of one target service text message in each target service text message and a third text feature of each candidate text message in the preset candidate text message set; and screening out at least one synonymy text message which is in synonymy association with the target service text message from the preset candidate text message set according to the determined similarity.

5. The method of claim 1, wherein the derived service text information comprises upper text information of the target service text information;

the determining at least one derivative service text message associated with each target service text message according to the first text feature of each target service text message respectively includes:

and screening at least one upper text message associated with each target service text message from a preset candidate text message set according to the first text feature of each target service text message and a preset matching rule.

6. The method according to any one of claims 1 to 5, wherein acquiring text information of each target service for a search service scenario comprises:

obtaining each service text information historically input by a plurality of objects in a search client, and taking the obtained each service text information as target service text information aiming at the search service scene; or

And acquiring each service text message from at least one database as target service text message aiming at the search service scene.

7. The method of any one of claims 1 to 5, further comprising:

responding to a search instruction triggered by a target object, and acquiring search service text information contained in the search instruction;

determining the search service text information and at least one service text information having a correlation with the search service text information from the service knowledge map according to the constructed service knowledge map; taking the search service text information and at least one piece of service text information which has a correlation with the search service text information as service text information to be recommended;

and recommending the determined page display information corresponding to the text information of the service to be recommended to the target object.

8. A service knowledge graph building apparatus, comprising:

the acquisition unit is used for acquiring text information of each target service aiming at a search service scene;

the determining unit is used for determining at least one derivative service text message respectively associated with each target service text message according to the first text feature of each target service text message;

and the processing unit is used for constructing a corresponding service knowledge graph according to the association relation between each target service text message and each associated at least one derivative service text message.

9. An electronic device, comprising a processor and a memory, wherein the memory stores program code which, when executed by the processor, causes the processor to perform the steps of the method of any of claims 1 to 7.

10. A computer-readable storage medium, characterized in that it comprises program code for causing an electronic device to carry out the steps of the method according to any one of claims 1 to 7, when said program code is run on said electronic device.

Images