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

Abstract

The application provides a service knowledge graph construction method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence. In the search service scenario, the application embodiment acquires target service text information; according to the first text features of each target service text information, at least one derivative service text information associated with each target service text information is determined; and according to the association relationship between each target service text information and the at least one derivative service text information associated therewith, a corresponding service knowledge graph is constructed. The service knowledge graph construction scheme provided by the application embodiment is not limited to the target service text information obtained, and the derivative service text information associated with the target service text information can be obtained, so that the service knowledge graph constructed contains more abundant information.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, and particularly relates to a service knowledge graph construction method and device, electronic equipment and storage medium. BACKGROUND

[0002] The knowledge graph refers to a tree topology network taking entities and concepts as nodes and taking relationships between the nodes as edges. The knowledge graph makes knowledge acquisition more direct, thereby realizing the convenience, intelligence and humanization of data search.

[0003] When searching information, information associated with a search keyword input by a user can be searched according to a pre-constructed knowledge graph. At present, the knowledge graph is constructed in an artificial construction manner. A large number of keywords are artificially obtained, and the obtained keywords are taken as nodes of the knowledge graph. Relationships between two nodes set artificially are taken as edges of the knowledge graph, and two nodes having an associated relationship are connected as edges of the knowledge graph. The construction of the knowledge graph is completed after the nodes and edges are generated. However, the knowledge graph obtained by using the artificial construction manner has great limitations in the quantity and types of information contained, and the constructed knowledge graph is too single. SUMMARY

[0004] The present application provides a service knowledge graph construction method and device, electronic equipment and storage medium, to flexibly construct a service knowledge graph with rich information.

[0005] In a first aspect, the present application provides a service knowledge graph construction method, comprising:

[0006] For a search service scenario, obtaining each target service text information;

[0007] According to a first text feature of each target service text information, determining at least one derivative service text information associated with each target service text information, respectively;

[0008] 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, constructing a corresponding service knowledge graph.

[0009] In a second aspect, the present application provides a service knowledge graph construction device, comprising:

[0010] An obtaining unit, configured to obtain each target service text information for a search service scenario;

[0011] A determining unit, configured to determine at least one derivative service text information associated with each target service text information according to a first text feature of each target service text information, respectively;

[0012] a processing unit, configured to construct a corresponding service knowledge graph according to the association relationship between the respective target service text information and the at least one derived service text information associated therewith.

[0013] Optionally, the processing unit is specifically configured to:

[0014] the following operations are performed respectively for the respective target service text information:

[0015] triplet information containing the target service text information, the association type and the derived service text information is generated respectively according to one target service text information and the at least one derived service text information associated therewith in the respective service text information; the number of the triplet information is the same as the number of the derived service text information associated with the one 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 triplet information.

[0016] Optionally, the determination unit is further configured to:

[0017] the service type to which each of the respective target service text information belongs is determined respectively according to the second text feature of the respective target service text information, and the service type to which each of the respective derived service text information belongs is determined respectively according to the third text feature of the respective derived service text information obtained;

[0018] the processing unit is further configured to:

[0019] after the corresponding service knowledge graph is constructed, in response to an instruction of displaying the service knowledge graph, the corresponding nodes are generated respectively according to the target service text information and the derived service text information contained in the respective triplet information generated, and the generated corresponding nodes are classified and displayed based on the service type corresponding to each of the respective target service text information and the service type corresponding to each of the respective derived service text information; and the edges between the respective nodes are generated according to the association type contained in the respective triplet information generated; and the constructed service knowledge graph is displayed according to the generated nodes and the edges between the nodes.

[0020] Optionally, the derived service text information includes synonymous text information in a synonymous association relationship with the target service text information.

[0021] the determination unit is further configured to:

[0022] Before the determining the at least one derived service text information respectively associated with the respective target service text information, respectively determining text content information respectively corresponding to the respective target service text information; respectively performing fusion processing on information content of the text content information respectively corresponding to the respective target service text information, to obtain a first text feature of the corresponding target service text information; and respectively determining text content information respectively corresponding to each candidate text information in a preset candidate text information set; respectively performing fusion processing on information content of the text content information respectively corresponding to the respective candidate text information, to obtain a third text feature of the corresponding candidate text information; wherein the text content information includes any one or combination of an edit distance, a semantic distance, co-occurrence information, and attribute information.

[0023] The determining unit is specifically configured to:

[0024] For each target service text information, the following operations are performed: based on the trained text relevance model, according to the first text feature of one target service text information in the target service text information and the third text feature of each candidate text information in the preset candidate text information set, determining the similarity between the one target service text information and each candidate text information in the preset candidate text information set; and according to the determined similarities, screening at least one synonymous text information in a synonymous association relationship with the one target service text information from the preset candidate text information set.

[0025] Optionally, the derived service text information includes superordinate text information of the target service text information.

[0026] The processing unit is specifically configured to:

[0027] According to the first text feature of each target service text information and a preset matching rule, screening at least one superordinate text information respectively associated with each target service text information from the preset candidate text information set.

[0028] Optionally, the obtaining unit is specifically configured to:

[0029] 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 for the search service scenario; or

[0030] Obtaining each service text information from at least one database as target service text information for the search service scenario.

[0031] Optionally, the processing unit is further configured to:

[0032] In response to a search instruction triggered by a target object, search service text information contained in the search instruction is acquired; according to the constructed service knowledge graph, the search service text information and at least one service text information having an association relationship with the search service text information are determined from the service knowledge graph; and the search service text information and the at least one service text information having an association relationship with the search service text information are taken as to-be-recommended service text information; and page display information corresponding to the to-be-recommended service text information is recommended to the target object.

[0033] In a third aspect, an embodiment of the present application provides an electronic device, comprising:

[0034] at least one processor; and

[0035] a memory in communication with the at least one processor; wherein

[0036] 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 perform the service knowledge graph construction method provided by the present application.

[0037] In a fourth aspect, an embodiment of the present application provides a computer readable medium storing computer executable instructions, and the computer executable instructions are used to perform the service knowledge graph construction method provided by the present application.

[0038] The present application has the following beneficial effects:

[0039] The service knowledge graph construction scheme provided by the embodiments of the present application, when constructing a service knowledge graph, first acquires a plurality of target service text information, and then determines at least one derivative service text information associated with each target service text information according to the text features of the target service text information. Therefore, at least one derivative service text information associated with a target service text information can be obtained, 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 therewith. Therefore, the service knowledge graph construction scheme provided by the embodiments of the present application is not limited to the target service text information obtained, and derivative service text information associated with the target service text information can be obtained, so that the service knowledge graph constructed contains more abundant information. BRIEF DESCRIPTION OF DRAWINGS

[0040] The accompanying drawings incorporated in and forming a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application without imposing on the application any undue limitations.

[0041] Figure 1 A service knowledge graph construction method provided by an embodiment of the present application is shown in the application scenario diagram.

[0042] Figure 2 A flowchart of a service knowledge graph construction method provided by an embodiment of the present application is shown in the flowchart.

[0043] Figure 3 A service text information determination method provided by an embodiment of the present application is shown in the schematic diagram.

[0044] Figure 4 Another service text information determination method provided by an embodiment of the present application is shown in the schematic diagram.

[0045] Figure 5 A service knowledge graph provided by an embodiment of the present application is shown in the schematic diagram.

[0046] Figure 6 An application scenario diagram of a service knowledge graph provided by an embodiment of the present application is shown in the application scenario diagram.

[0047] Figure 7 A search method based on a service knowledge graph provided by an embodiment of the present application is shown in the flowchart.

[0048] Figure 8 A service knowledge graph construction device provided by an embodiment of the present application is shown in the structural diagram.

[0049] Figure 9 An electronic device provided by an embodiment of the present application is shown in the structural diagram.

[0050] Figure 10 A computing device provided by an embodiment of the present application is shown in the structural diagram. DETAILED DESCRIPTION

[0051] In order to make the 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 described clearly and completely below with reference to the drawings.

[0052] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0053] Hereinafter, part of the terms in the embodiments of the present application are explained and described to facilitate the understanding of those skilled in the art.

[0054] 1. Vertical search: Vertical search is a professional search engine for a certain industry, which is a subdivision and extension of search engines. It is a kind of integration of certain special information in the library, and the required data is extracted and processed in a certain form and returned to the user. For example, public number search, applet search, etc.

[0055] 2. Service search: Users can search for services through search engines. The user inputs the search keyword (query), and the web page or client can directly display the service that meets the user's search keyword to the user. For example, when the user searches for a nanny, the service search can directly provide the service menu of hiring a nanny. Service search is a kind of vertical search.

[0056] 3. Knowledge graph: It is a theory that combines the theories and methods of mathematics, graphics, information visualization technology, information science, and citation analysis, co-occurrence analysis, and uses visual graph to display the core structure, development history, frontier field and overall knowledge architecture of the subject to achieve the purpose of multi-disciplinary integration. The complex knowledge field is displayed through data mining, information processing, knowledge measurement and graphics drawing, revealing the dynamic development law of the knowledge field, providing practical and valuable reference for subject research. Through effective processing, processing and integration of complex data documents, it is transformed into simple and clear "entity-relation-entity" triples, and a large number of knowledge topology graphs are aggregated.

[0057] 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 people and computers using natural language. Natural language processing is a science that integrates linguistics, computer science and mathematics.

[0058] 5. Semantics: The information contained in language is semantics. Symbols are the carriers of language, and symbols themselves have no meaning. Only symbols with meaning can be used, at which time language is converted into information, and the meaning of language is semantics (Semantic).

[0059] 6、Terminal: also known as User Equipment (UE), Mobile Station (MS), Mobile Terminal (MT), etc., is a device that provides voice and / or data connectivity to users, such as handheld devices with wireless connectivity, vehicle-mounted devices, etc. Currently, some examples of terminals are: mobile phones, tablet computers, notebook computers, palmtop computers, Mobile Internet Devices (MIDs).

[0060] 7、Client: can refer to both software-based application programs (APPs) and terminal devices. It has a visual display interface and can interact with users; it provides local services for clients. For software-based application programs, in addition to some applications that only run locally, they are generally installed on ordinary client terminals and need to run in cooperation with servers. After the development of the Internet, commonly used application programs include email clients when sending and receiving emails, and instant messaging clients, etc. For this type of application program, there needs to be a corresponding server and service program in the network to provide corresponding services, such as database services, configuration parameter services, etc., so that specific communication connections need to be established between the client terminal and the server to ensure the normal operation of the application program.

[0061] 8、Server: can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, etc. Basic cloud computing services.

[0062] The design idea of the embodiments of the present application is briefly introduced as follows:

[0063] Artificial Intelligence (AI) is the use of digital computers or computer-controlled machines 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 technology of computer science, which aims to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making.

[0064] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software level technology. Artificial intelligence basic technology generally includes, such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics and other technologies. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning and other major directions.

[0065] Machine learning (ML) is a study of how computers simulate or implement human learning behavior to acquire new knowledge or skills, and reorganize existing knowledge structure to continuously improve their performance. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and rule-based learning. Machine learning is the core of artificial intelligence and the fundamental approach to making computers intelligent, and its applications are widespread in various fields of artificial intelligence.

[0066] With the continuous popularity of search engines and various clients, more and more users can use search functions in various scenarios. The user inputs a search keyword in the search engine or the client, and the search engine or the client can display information related to the user input search keyword to the user based on a pre-constructed knowledge graph. At present, the knowledge graph is constructed by using an artificial construction method. A large number of keywords are obtained by manual acquisition, and the obtained keywords are used as nodes of the knowledge graph. The relationship between two nodes set by manual setting is used to connect two nodes having a correlation relationship as an edge of the knowledge graph. The construction of the knowledge graph is completed after the nodes and edges are generated. However, the knowledge graph obtained by using the artificial construction method has great limitations in the number and types of information contained. Moreover, the knowledge graph constructed at present is for an article or commodity search scenario. Based on the constructed knowledge graph, other articles or commodities similar to the articles or commodities searched by the user are recommended to the user. For example, the user inputs a search keyword "air conditioner", and various brands and functions of air conditioners are recommended to the user. At present, there is no service knowledge graph constructed for a service scenario.

[0067] Therefore, this application provides a method for constructing a service knowledge graph. For a search service scenario, it acquires text information of various target services; determines at least one derivative service text information associated with each target service text information based on its first text feature; and constructs a corresponding service knowledge graph based on the association between each target service text information and its associated at least one derivative service text information. The service knowledge graph construction scheme provided in this application first acquires multiple target service text information, then determines at least one derivative service text information associated with each target service text information based on its text features. Therefore, at least one associated derivative service text can be obtained from a single target service text information, thus enriching the service text information. Furthermore, after obtaining at least one derivative service text information associated with each target service text information, a corresponding service knowledge graph is constructed based on the association between each target service text information and its associated at least one derivative service text information. Therefore, the service knowledge graph construction scheme provided in this application is not limited to the acquired target service text information; associated derivative service text information can be obtained from the target service text information, resulting in a more comprehensive service knowledge graph.

[0068] After introducing the design concept of the embodiments of this application, the following is a brief introduction to the application scenarios to which the technical solutions of the embodiments of this application can be applied. It should be noted that the application scenarios described below are only for illustrating the embodiments of this application and are not intended to limit the scope. In specific implementation, the technical solutions provided by the embodiments of this application can be flexibly applied according to actual needs.

[0069] like Figure 1 The diagram shown is an exemplary application scenario of this application embodiment, including user 10, terminal 11, and server 12.

[0070] One possible application scenario is that user 10 triggers a service knowledge graph construction command through the display interface of terminal 11, and terminal 11 sends the construction command triggered by user 10 to server 12; after receiving the construction command, server 12 obtains the text information of each target service for the search service scenario; based on the first text feature of each target service text information, it determines at least one derivative service text information associated with each target service text information; based on the association relationship between each target service text information and its associated at least one derivative service text information, it constructs the corresponding service knowledge graph; after constructing the service knowledge graph, server 12 stores the constructed service knowledge graph.

[0071] Wherein, the server of the embodiment of the application can be a stand-alone server, or can be a server cluster composed of multiple servers. The terminal 11 and the server 12 are connected through a network, which can be a local area network, a wide area network, etc.

[0072] The service knowledge graph construction method provided by the exemplary embodiments of the application will be described below in combination with the application scenarios described above, with reference to Figures 2-5 It should be noted that the above application scenarios are only shown for the purpose of facilitating the understanding of the spirit and principles of the application, and the embodiments of the application are not limited in this respect. On the contrary, the embodiments of the application can be applied to any applicable scenario.

[0073] As Figure 2 shown, a flowchart of a service knowledge graph construction method provided by the embodiment of the application, which can include the following steps:

[0074] Step S201, for a search service scenario, obtaining each target service text information;

[0075] Step S202, determining at least one derived service text information associated with each target service text information according to the first text features of each target service text information, respectively;

[0076] Step S203, constructing a corresponding service knowledge graph according to the association relationship between each target service text information and the at least one derived service text information associated therewith.

[0077] The target service text information of the embodiment of the application can be a keyword with a service search intent;

[0078] The acquisition method of the target service text information of the embodiment of the application can include but is not limited to the following methods:

[0079] Method 1, obtaining each service text information historically input by multiple objects in a search client, and taking each service text information obtained as the target service text information for the search service scenario;

[0080] In this way, a large amount of service text information historically input by users in a search client can be collected; since the service text information historically input by users in a search client is generally a keyword with a service search intent, a large amount of service text information historically input by users in a search client can be taken as the target service text information for the search service scenario.

[0081] For example, the service text information historically input by users in a search client can be "repairing air conditioners", "repairing washing machines", "dry cleaning", "finding a nanny", "finding a job", etc.

[0082] Mode 2, obtaining each service text information from at least one database as target service text information for searching a service scenario;

[0083] In this mode, the embodiment of the application can obtain service text information from one or more databases when constructing a service knowledge graph. The service text information obtained from one or more databases can be text information with a service search intent.

[0084] When constructing a service knowledge graph, the embodiment of the application needs to classify nodes in the service knowledge graph according to the category system to which the nodes belong. After obtaining each target service text information, the service type to which each target service text information belongs is determined.

[0085] 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.

[0086] In some embodiments, the service type to which each target service text information belongs is determined according to the following mode:

[0087] Based on the trained text classification model, the service type to which each target service text information belongs is determined.

[0088] In implementation, each target service text information can be input into the trained text classification model, and the output of the trained text classification model can be used as the service type to which each target service text information belongs.

[0089] The trained text classification model includes a feature extraction network and a classification network.

[0090] The feature extraction network of the trained text classification model is randomly initialized by embedding the target service text information.

[0091] The classification network of the trained text classification model classifies the embedding vector obtained by random initialization to obtain the service type to which each target service text information belongs.

[0092] The mode of obtaining the service type to which each target service text information belongs based on the trained text classification model is described in detail as follows:

[0093] Based on the feature extraction network of the trained text classification model, the following operations are performed for each target service text information:

[0094] extract a character vector, a word vector and a position vector representing the relative position relationship between characters in a target service text information from each target service text information; and perform fusion processing on the extracted character vector, word vector and position vector to obtain a second text feature of the target service text information.

[0095] After obtaining the second text feature of the target service text information based on the feature extraction network of the trained text classification model, the classification network of the trained text classification model is used to determine the service type to which each target service text information belongs.

[0096] The classification network of the trained text classification model performs the following operations on each target service text information:

[0097] The second text feature of each target service text information is subjected to mean value processing to obtain a classification result representing the probability of each target service text information belonging to each preset service type; and the service type to which each target service text information belongs is determined according to the classification result.

[0098] It should be noted that the text classification model of the embodiment of the application can be a Fasttext model.

[0099] After obtaining each target service text information, the embodiment of the application can also determine the attribute information of each target service text information.

[0100] In determining the attribute information of each target service text information, the embodiment of the application obtains the attribute feature corresponding to each target service text information.

[0101] The attribute feature corresponding to the target service text information includes but is not limited to a part-of-speech feature, an embedding feature and a word frequency feature.

[0102] For the part-of-speech feature, the part-of-speech feature of each target service text information can be obtained by using the Jieba word segmentation; for example, the part-of-speech feature can be a noun, a verb, a gerund, etc.

[0103] For the embedding feature, the trained word2vec model can be used to obtain the embedding feature of each target service text information.

[0104] For the word frequency feature, the trained TF IDF model can be used to obtain the word frequency feature of each target service text information.

[0105] After obtaining the attribute features corresponding to the target service text information, 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.

[0106] After obtaining each target service text information, the embodiment of the present application determines at least one derived service text information associated with each target service text information according to the first text features of each target service text information.

[0107] It should be noted that each target service text information in the embodiment of the present application can determine at least one derived service text information.

[0108] The derived service text information includes synonymous text information in a synonymous association relationship with the target service text information, and superordinate text information of the target service text information.

[0109] The determination methods of the two types of derived service text information will be described below:

[0110] 1. The derived service text information is synonymous text information in a synonymous association relationship with the target service text information.

[0111] For example, when the target service text information is an entity word, the derived service text information is a synonym or an extended word of the entity word.

[0112] In determining the synonymous text information of the target service text information, the first text features of each target service text information can be determined based on a trained text correlation model.

[0113] Based on the trained text correlation model, the first text features of each target service text information are extracted.

[0114] In the implementation, for each target service text information, the first text features of each target service text information can be obtained according to the following methods:

[0115] The text content information corresponding to each target service text information is determined respectively, and the information content of the text content information corresponding to each target service text information is fused respectively to obtain the first text features of the corresponding target service text information.

[0116] In addition, when the text correlation model is trained, at least one synonymous text information in a synonymous association relationship with the target service text information is selected from the candidate text information set according to the first text features of each target service text information and the third text features of each candidate text information in the candidate text information set.

[0117] In an 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 according to the following manner:

[0118] determining the respective corresponding text content information of each candidate text information in the preset candidate text information set; and performing fusion processing on the information content of the respective corresponding text content information of each candidate text information, to obtain the third text feature of the corresponding candidate text information.

[0119] It should be noted that the text content information includes any one or combination of an edit distance, a semantic distance, co-occurrence information, and attribute information.

[0120] The edit distance is a quantitative measure of the degree of difference between two strings in text information, and the measure is how many times of processing are needed to change one string into another string. The edit distance can be used in NLP. The semantic distance is the distance in the semantic space. The co-occurrence information is the phenomenon that the information described by the text information commonly occurs.

[0121] The determination manner of the attribute information can be referred to the description above. For example, when the target service text information includes a service word, the attribute information includes a service entity word, a service behavior word, a service state word, and a service compound word, etc.

[0122] 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, at least one derived service text information associated with each target service text information is obtained based on the trained text relevance model.

[0123] For example, the text relevance model of the embodiments of the present application can be a linear LR model.

[0124] In an implementation, the following operations are performed for each target service text information:

[0125] determining the similarity between one target service text information and each candidate text information in the preset candidate text information set based on the trained text relevance model, according to the first text feature of the one target service text information and the third text feature of each candidate text information in the preset candidate text information set; and selecting at least one synonymous text information in synonymous association relationship with the one target service text information from the preset candidate text information set according to the determined similarities.

[0126] It should be noted that in determining the synonymous text information associated with the target text information, the 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.

[0127] For example, when the target text information is "air conditioner", the corresponding synonymous text information includes synonymous words and extension words; wherein the synonymous word corresponding to "air conditioner" obtained can be "refrigeration air conditioner", and the extension word corresponding to "air conditioner" obtained can be car air conditioner, central air conditioner, air conditioner equipment, air conditioner, air conditioner vehicle, air conditioner network, air, variable frequency air conditioner, etc.

[0128] 2, the derived service text information and the target service text information are the superordinate text information of the target service text information;

[0129] For example, when the target service text information is an entity word, the derived service text information is the superordinate word of the entity word.

[0130] In some embodiments, at least one superordinate text information associated with each target service text information is obtained according to the following manner:

[0131] According to the first text feature of each target service text information and the preset matching rule, at least one superordinate text information associated with each target service text information is screened out from the preset candidate text information set.

[0132] It should be noted that in obtaining the associated superordinate text information according to the target service text information, the application embodiment can obtain the superordinate text information set associated with the target service text information based on the artificial pattern mining manner. For example, pattern mining determines that A belongs to B, and B is the superordinate text information of A.

[0133] Through the above described manner, 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 superordinate text information of the target service text information can be obtained.

[0134] For example, as shown in the table Figure 3 of the application embodiment, the target service text information includes air conditioner: it is determined that the service type to which "air conditioner" belongs is home service; it is determined that the attribute information of "air conditioner" is service entity word; it is determined that the extension words of "air conditioner" include car air conditioner, central air conditioner, air conditioner equipment, air conditioner, air conditioner vehicle, air conditioner network, air, variable frequency air conditioner; it is determined that the synonymous word of "air conditioner" is refrigeration air conditioner; it is determined that the superordinate word of "air conditioner" includes household electrical appliances, home electrical appliances, home appliances, electrical appliances, consumer electronics.

[0135] As shown in the table Figure 4As shown, the target service text information in the embodiment of the present application includes "quickly": determine the service type to which "quickly" belongs as housekeeping or transportation; under the housekeeping type, determine the attribute information of "quickly" as a service state word, and determine the extension words of "quickly" to include fast, speed; under the transportation type, determine the attribute information of "quickly" as a service state word, and determine the extension words of "quickly" to include expressway, speed, and highway.

[0136] After determining the at least one derived service text information associated with each target service text information, the embodiment of the present application constructs a corresponding service knowledge graph according to the association relationship between each target service text information and the at least one derived service text information associated therewith;

[0137] In some embodiments, the service knowledge graph can be constructed according to the following manner:

[0138] According to one target service text information in each service text information and the at least one derived service text information associated therewith, generate a triple information containing the target service text information, the association type, and the derived service text information; wherein the number of triple information is the same as the number of derived service text information associated with one target service text information, and the association type is the type of association relationship between the target service text information and the derived service text information contained in the triple information.

[0139] It should be noted that the service knowledge graph constructed by the embodiment of the present application can be stored in the form of triple information when stored, and each triple information is in the form of <entity-relation-entity>; wherein the entity can be a target service text information or a derived service text information, and the relation is the association relationship between the two entities;

[0140] Among them, the association relationship between the two entities in the embodiment of the present application includes but is not limited to: belongs to, acts on, behavior is, synonymous, is located, brand is, main service is, service area, service population, and occurrence.

[0141] For example, the target service text information is "washing machine repair", the determined superordinate text information includes home appliance repair, and the determined synonymous text information includes washing machine and repair; the target service text information is "toilet repair", the determined superordinate text information includes house repair and kitchen and bathroom repair, and the determined synonymous text information includes toilet and repair; the target service text information is "kitchen and bathroom repair", the determined superordinate text information includes house repair, and the determined synonymous text information includes kitchen and bathroom appliance, repair, and repair;

[0142] The generated triple information includes:

[0143] <washer repair - belongs to - appliance repair>, <washer repair - synonymous - washer>, <washer repair - synonymous - repair>, <toilet repair - belongs to - house repair>, <toilet repair - belongs to - kitchen and bathroom repair>, <toilet repair - synonymous - toilet>, <toilet repair - synonymous - repair>, <kitchen and bathroom repair - belongs to - house repair>, <kitchen and bathroom repair - synonymous - kitchen and bathroom appliances>, <kitchen and bathroom repair - synonymous - repair>, <kitchen and bathroom repair - synonymous - repair>.

[0144] In addition, the embodiments of the present application also need to determine the service type of each target service text information and derived service text information. The determination method of the service type to which the target service text information belongs can be referred to the description above. The determination method of the service type to which the derived service text information belongs is similar to the determination method of the service type to which the target service type belongs, which will not be described in detail here.

[0145] When the constructed service knowledge graph is displayed, the service knowledge graph can be displayed in a tree topology structure. An optional embodiment is that when the constructed service knowledge graph is displayed, it is displayed according to different service types.

[0146] In the implementation, the service type to which each target service text information belongs is determined respectively, and the service type to which each derived service text information belongs is determined respectively.

[0147] In response to the instruction of 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 each triple information, and the generated corresponding nodes are displayed according to the service type corresponding to each target service text information and the service type corresponding to each derived service text information. According to the association type contained in the generated each triple information, the edges between the nodes are generated. According to the generated nodes and the edges between the nodes, the constructed service knowledge graph is displayed.

[0148] For example, the service category system under the housekeeping service type includes appliance repair and house repair. The triple information under the appliance repair service category system is as follows:

[0149] <washer repair - synonymous - washer>, <washer repair - upper - repair>, <washer - upper - appliance>, <appliance - synonymous - household goods>;

[0150] The triple information under the house repair service category system is as follows:

[0151] <toilet repair-synonym-toilet> <toilet repair-hypernym-plumbing repair> <toilet repair-hypernym-repair> <toilet-hypernym-plumbing fixture> <plumbing repair-hypernym-plumbing fixture> <plumbing repair-synonym-repair> <plumbing repair-synonym-repair> <plumbing fixture-hypernym-household item> <repair-synonym-repair> <repair-synonym-cleaning> <repair-synonym-cleaning>.

[0152] The service category system included under the government service type includes traffic management services and household administration services. The triple information included under the traffic management service category system is as follows:

[0153] <traffic violation-synonym-traffic violation query> <traffic violation query-hypernym-traffic violation> <traffic violation query-hypernym-query> <traffic violation-hypernym-traffic> <traffic violation-synonym-violation>.

[0154] The triple information included under the household administration service category system is as follows:

[0155] <id card-synonym-id card loss> <id card-synonym-query> <id card loss-hypernym-id card> <id card loss-hypernym-loss>.

[0156] Based on the above triple information, the generated service knowledge graph is as shown in Figure 5 .

[0157] After constructing the service knowledge graph, the constructed service knowledge graph can be applied to a search scenario. An optional application scenario diagram of the service knowledge graph is as shown in Figure 6 , which includes a user 60, a terminal 61, and a server 62. The terminal 61 has a client installed thereon.

[0158] The user 60 triggers a search instruction through the client installed on the terminal 61, and the client obtains search service text information contained in the search instruction. The search service text information is sent to the server 62. The server 62 determines the search service text information and at least one service text information associated with the search service text information from the constructed service knowledge graph according to the service knowledge graph. The search service text information and at least one service text information associated with the search service text information are taken as recommended service text information. The page display information corresponding to the determined recommended service text information is returned to the client, and the client displays the page display information corresponding to the recommended service text information to the target object in the display interface.

[0159] As shown in Figure 7 , a search method based on the service knowledge graph provided by the embodiment of the application is shown in the flowchart. The method can include the following steps:

[0160] Step S701: The client responds to the search command triggered by the target object and obtains the search service text information contained in the search command;

[0161] Step S702: The client sends the search service text information to the server;

[0162] Step S703: The server determines the search service text information and at least one service text information that is related to the search service text information from the constructed service knowledge graph.

[0163] Step S704: The server uses the search service text information and at least one service text information that is related to the search service text information as the service text information to be recommended.

[0164] Step S705: The server returns the page parameters corresponding to the determined text information of the service to be recommended to the client;

[0165] Step S706: The client recommends the page display information corresponding to the text information of the service to be recommended to the target object.

[0166] Based on the same inventive concept, this application also provides a title text processing device. Since the principle of this device in solving the problem is similar to the title text processing method described above, the implementation of this device can refer to the implementation of the method, and the repeated parts will not be described again.

[0167] like Figure 8 The diagram shown is a structural schematic of a service knowledge graph construction device provided in an embodiment of this application, comprising:

[0168] The acquisition unit 801 is used to acquire text information of each target service for search service scenarios;

[0169] The determining unit 802 is used to determine at least one derivative service text information associated with each of the target service text information based on the first text features of each target service text information.

[0170] The processing unit 803 is used to construct a corresponding service knowledge graph based on the association relationship between each target service text information and at least one associated derivative service text information.

[0171] Optionally, the processing unit 803 is specifically used for:

[0172] Perform the following operations on each of the target service text information:

[0173] According to one target service text information and at least one derived service text information associated with the target service text information in the respective service text information, a triple information containing the target service text information, an association type and the derived service text information is generated respectively; the number of the triple information is the same as the number of the derived service text information associated with the one 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.

[0174] Optionally, the determination unit 802 is further configured to:

[0175] According to the second text features of the respective target service text information, the service type to which each of the respective target service text information belongs is determined; and according to the third text features of the respective derived service text information obtained, the service type to which each of the respective derived service text information belongs is determined.

[0176] The processing unit 803 is further configured to:

[0177] After the corresponding service knowledge graph is constructed, in response to an instruction of displaying the service knowledge graph, the corresponding nodes are respectively generated according to the target service text information and the derived service text information contained in the respective triple information generated, and the generated corresponding nodes are classified and displayed based on the service type corresponding to each of the respective target service text information and the service type corresponding to each of the respective derived service text information; and the edges between the respective nodes are generated according to the association type contained in the respective triple information generated; and the constructed service knowledge graph is displayed according to the generated nodes and the edges between the nodes.

[0178] Optionally, the derived service text information includes synonymous text information in a synonymous association relationship with the target service text information.

[0179] The determination unit 802 is further configured to:

[0180] Before the at least one derived service text information associated with each of the respective target service text information is determined, the text content information corresponding to each of the respective target service text information is determined respectively; the information content of the text content information corresponding to each of the respective service text information is fused respectively to obtain the first text feature of the corresponding target service text information; and the text content information corresponding to each of the respective candidate text information in the preset candidate text information set is determined; the information content of the text content information corresponding to each of the respective candidate text information is fused respectively to obtain the third text feature of the corresponding candidate text information; wherein the text content information includes any one or combination of an edit distance, a semantic distance, co-occurrence information and attribute information.

[0181] The determination unit 802 is specifically configured to:

[0182] For each target service text information, the following operations are performed: based on the trained text relevance model, according to the first text feature of one target service text information in the 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 one target service text information and each candidate text information in the preset candidate text information set is determined; and according to the determined similarities, at least one synonymous text information in a synonymous association relationship with the one target service text information is screened out from the preset candidate text information set.

[0183] Optionally, the derived service text information includes the superordinate text information of the target service text information.

[0184] The determination unit 802 is specifically configured to:

[0185] According to the first text feature of each target service text information and a preset matching rule, at least one superordinate text information associated with each target service text information is screened out from the preset candidate text information set.

[0186] Optionally, the acquisition unit 801 is specifically configured to:

[0187] Acquire each service text information historically input by a plurality of objects in a search client, and acquire the each service text information as target service text information for the search service scenario; or

[0188] Acquire each service text information from at least one database as target service text information for the search service scenario.

[0189] Optionally, the processing unit 803 is further configured to:

[0190] In response to a search instruction triggered by a target object, acquire search service text information contained in the search instruction; according to the constructed service knowledge graph, determine the search service text information and at least one service text information in an association relationship with the search service text information from the service knowledge graph; and take the search service text information and the at least one service text information in an association relationship with the search service text information as to-be-recommended service text information; and recommend page display information corresponding to the to-be-recommended service text information to the target object.

[0191] For ease of description, the above parts are divided into modules (or units) according to functions and are described respectively. Of course, functions of the modules (or units) can be implemented in one or more software or hardware in the implementation of the present application.

[0192] Those skilled in the art can understand that each aspect of the present application can be implemented as a system, a method or a program product. Therefore, each aspect of the present application can be specifically implemented as follows: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which can be collectively referred to as "circuitry", "module" or "system" here.

[0193] In some possible implementation manners, the embodiments of the present application further provide an electronic device, as shown in Figure 9 , the electronic device 900 can at least include at least one processor 901 and at least one memory 902. The memory 902 stores program code, when the program code is executed by the processor 901, the processor 901 executes the steps in the service knowledge graph construction method according to various exemplary embodiments of the present application described in the present specification, for example, the processor 901 can execute the steps as shown in Figure 2 or as shown in Figure 7 .

[0194] In some possible implementation manners, the embodiments of the present application further provide a computing device, which can at least include at least one processing unit and at least one storage unit. The storage unit stores program code, when the program code is 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 in the present specification, for example, the processor 901 can execute the steps as shown in Figure 2 or as shown in Figure 7 .

[0195] The computing device 1000 according to this embodiment of the present application will be described below with reference to Figure 10 . Figure 10 The computing device 1000 is only an example and should not bring any limitation to the functions and use range of the embodiments of the present application.

[0196] As shown in Figure 10 , the computing device 1000 is in the form of a general computing device. The components of the computing device 1000 can 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).

[0197] Bus 1003 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a local bus using any of a variety of bus architectures.

[0198] Storage 1002 can also include a read only memory (ROM) 1023 in the form of flash, among others.

[0199] Storage 1002 can also include a program / utility 1025 having a set of program modules 1024 such as an operating system, one or more application programs, other program modules, and program data, each of which can govern, among other things, the operation of the network environment.

[0200] Computing device 1000 can also communicate with one or more external devices 1004 such as a keyboard or pointing device, using I / O interface 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000, such as a display, a speaker, a headphone, a microphone, a printer, and so forth. One or more devices that enable a user to interact with computing device 1000 can be referred to as human interface devices (HIDs) 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000 in a wireless manner, such as a personal digital assistant, a cellular telephone, a handset, a Bluetooth® device and so forth. One or more devices that enable a user to interact with computing device 1000 in a wireless manner can be referred to as wireless HIDs 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000 in a wired manner, such as a personal digital assistant, a cellular telephone, a handset, a Bluetooth® device and so forth. One or more devices that enable a user to interact with computing device 1000 in a wired manner can be referred to as wired HIDs 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000 in a virtual manner, such as a personal digital assistant, a cellular telephone, a handset, a Bluetooth® device and so forth. One or more devices that enable a user to interact with computing device 1000 in a virtual manner can be referred to as virtual HIDs 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000 in a remote manner, such as a personal digital assistant, a cellular telephone, a handset, a Bluetooth® device and so forth. One or more devices that enable a user to interact with computing device 1000 in a remote manner can be referred to as remote HIDs 1005. I / O interface 1005 can also interface to one or more devices that enable a user to interact with computing device 1000 in a networked manner, such as a personal digital assistant, a cellular telephone, a handset, a Bluetooth® device and so forth. One or more devices that enable a user to interact with computing device 1000 in a networked manner can be referred to as networked HIDs 1005.

[0201] In some possible implementation, 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 codes for causing a computer device to perform the steps of the service knowledge graph construction method according to various exemplary embodiments of the present application described above in the specification when the program product is run on the computer device, for example, the computer device can perform the steps as shown in Figure 2 or as shown in Figure 7 .

[0202] A program product can employ any combination of one or more computer-readable media. The computer-readable media can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include the following: an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0203] While the preferred embodiments of the application have been described, additional variations and modifications can be employed, as will be appreciated by those of ordinary skill in the art once advised of the essential inventive concepts. Therefore, the scope of the application should be determined with reference to the appended claims and equivalents thereof.

[0204] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and equivalents thereof, the application can be practiced otherwise than as specifically described.

Claims

1. A service knowledge graph construction method, characterized in that, The method comprises: For a search service scenario, 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 for the search service scenario; According to a first text feature of each target service text information, determining at least one derived service text information associated with each target service text information respectively; the derived service text information comprises at least one of the following contents: synonymous text information in a synonymous association relationship with the target service text information, and superordinate text information of the target service text information; the first text feature is determined based on at least attribute information of the corresponding target service text information; the attribute information comprises one or more of service entity words, service behavior words, service state words, and service compound words; Based on a service type to which the target service text information belongs, constructing a corresponding service knowledge graph according to an association relationship between each target service text information and at least one derived service text information associated therewith.

2. The method of claim 1, wherein, The construction of the corresponding service knowledge graph according to the association relationship between each target service text information and at least one derived service text information associated therewith comprises: For each target service text information, the following operations are performed respectively: According to one target service text information and at least one derived service text information associated therewith in each service text information, generating a triple information comprising the target service text information, an association type, and the derived service text information respectively; wherein the number of the triple information is the same as the number of the derived service text information associated with the one 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: According to a second text feature of each target service text information, determining a service type to which each target service text information belongs respectively; and according to a third text feature of each derived service text information obtained, determining a service type to which each derived service text information belongs respectively; After the construction of the corresponding service knowledge graph, the method further comprises: In response to an instruction to display the service knowledge graph, generating corresponding nodes according to the target service text information and the derived service text information contained in each generated triple information, and classifying and displaying the generated corresponding nodes based on the service type corresponding to each target service text information and the service type corresponding to each derived service text information; and Generating edges between nodes according to the association type contained in each generated triple information; Displaying the constructed service knowledge graph according to the generated nodes and the edges between the nodes.

4. The method of claim 1, wherein, The derived service text information comprises synonymous text information in a synonymous association relationship with the target service text information; Before the determination of at least one derived service text information associated with each target service text information, the method further comprises: determine respective text content information corresponding to each of the target service text information respectively; perform fusion processing on information content of the respective text content information corresponding to each of the target service text information respectively to obtain first text features of the corresponding target service text information; and determine respective text content information corresponding to each of the candidate text information in the preset candidate text information set; perform fusion processing on information content of the respective text content information corresponding to each of the candidate text information to obtain third text features of the corresponding candidate text information; wherein the text content information includes any one or combination of an edit distance, a semantic distance, co-occurrence information, and attribute information; The determining, according to the first text features of the target service text information respectively, of at least one derived service text information associated with each of the target service text information includes: For each target service text information, the following operations are performed: based on a trained text relevance model, determining a similarity between one target service text information and each of the candidate text information in the preset candidate text information set according to the first text feature of the one target service text information and the third text features of each of the candidate text information in the preset candidate text information set; and according to the determined similarities, screening at least one synonymous text information in a synonymous association relationship with the one target service text information from the preset candidate text information set.

5. The method of claim 1, wherein, The derived service text information includes superordinate text information of the target service text information. The determining, according to the first text features of the target service text information respectively, of at least one derived service text information associated with each of the target service text information includes: According to the first text features of each target service text information and a preset matching rule, screening at least one superordinate text information associated with each target service text information from a preset candidate text information set.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes: In response to a search instruction triggered by a target object, obtaining search service text information contained in the search instruction; According to the constructed service knowledge graph, determining the search service text information and at least one service text information in an association relationship with the search service text information from the service knowledge graph; and taking the search service text information and the at least one service text information in an association relationship with the search service text information as to-be-recommended service text information; Recommending page display information corresponding to the determined to-be-recommended service text information to the target object. 7.A service knowledge graph construction apparatus characterized by comprising: It includes: An obtaining unit is configured to, for a search service scenario, obtain each service text information historically input by a plurality of objects in a search client, and take the obtained each service text information as target service text information for the search service scenario. A determining unit is configured to determine at least one derived service text information associated with each of the target service text information according to a first text feature of each of the target service text information, wherein the derived service text information comprises at least one of the following: synonymous text information in a synonymous association relationship with the target service text information, and superordinate text information of the target service text information; the first text feature is determined based on at least attribute information of the corresponding target service text information; and the attribute information comprises one or more of a service entity word, a service behavior word, a service state word, and a service compound word. A processing unit is configured to construct a corresponding service knowledge graph according to an association relationship between each of the target service text information and the at least one derived service text information associated therewith based on a service type to which the target service text information belongs.

8. An electronic device, comprising: The device comprises 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 any of the methods of claims 1-6.

9. A computer-readable storage medium, characterized in that, The device comprises program code which, when executed on an electronic device, causes the electronic device to perform the steps of any of the methods of claims 1-6.

Images