CN112558913A

CN112558913A - Conversation method and device based on aggregated card, computer equipment and storage medium

Info

Publication number: CN112558913A
Application number: CN202011476665.1A
Authority: CN
Inventors: 赵凌燕
Original assignee: Ping An Life Insurance Company of China Ltd
Current assignee: Ping An Life Insurance Company of China Ltd
Priority date: 2020-12-15
Filing date: 2020-12-15
Publication date: 2021-03-26

Abstract

The application belongs to the technical field of artificial intelligence and provides a conversation method and device based on a polymerization card, computer equipment and a computer readable storage medium. According to the method and the device, voice input by a user is acquired, voice recognition is performed on the voice, a dialogue text corresponding to the voice is acquired, natural language processing is performed on the dialogue text, an entity and intention contained in the dialogue text are acquired, target display content corresponding to the intention is acquired according to the entity and the intention, a preset display control corresponding to the target display content is called according to the target display content, the target display content is displayed in a card form through the preset display control, the target display content is displayed in the card form, the target display content can be presented in a structured form, automatic assembly can be performed on the preset display control according to the target display content, dynamic aggregation of the card is achieved, and convenience in dialogue text display is improved.

Description

Conversation method and device based on aggregated card, computer equipment and storage medium

Technical Field

The present application relates to the field of artificial intelligence technologies, and in particular, to a conversation method and apparatus based on a syndication card, a computer device, and a computer-readable storage medium.

Background

A dialog system in natural language processing is a computer system capable of performing coherent dialog with a human, i.e., a dialog-based interaction (abbreviated as CUI). In the conventional technology, when a dialogue system is used for self-service, most answers of users are answered in a text form, and text answers corresponding to the questions are output generally according to the questions of the users. For example, for insurance products, when the insurance intelligent customer service robot answers the user questions through the dialogue system, the questions of the user are answered in the form of texts aiming at the product guarantee details included in the insurance products, so that too simple answers are not clear, and too long answer users cannot see the key points, thereby resulting in lower dialogue efficiency, reduced communication efficiency and waste of self-service resources.

Disclosure of Invention

The application provides a conversation method and device based on a syndication card, computer equipment and a computer readable storage medium, which can solve the problem of low self-service conversation efficiency in the traditional technology.

In a first aspect, the present application provides a syndication card-based conversation method, the method comprising: acquiring voice input by a user, and performing voice recognition on the voice to obtain a dialog text corresponding to the voice; carrying out natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text; according to the entity and the intention, target display content corresponding to the intention is obtained; calling a preset display control corresponding to the target display content according to the target display content; and displaying the target display content in a card form through the preset display control.

In a second aspect, the present application further provides a conversation device based on syndication cards, including: the device comprises a first acquisition unit, a second acquisition unit and a processing unit, wherein the first acquisition unit is used for acquiring voice input by a user and carrying out voice recognition on the voice to obtain a dialog text corresponding to the voice; the processing unit is used for carrying out natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text; the second acquisition unit is used for acquiring target display content corresponding to the intention according to the entity and the intention; the calling unit is used for calling a preset display control corresponding to the target display content according to the target display content; and the display unit is used for displaying the target display content in a card form through the preset display control.

In a third aspect, the present application further provides a computer device, which includes a memory and a processor, where the memory stores a computer program thereon, and the processor implements the steps of the aggregated card-based conversation method when executing the computer program.

In a fourth aspect, the present application further provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, causes the processor to perform the steps of the aggregated card-based conversation method.

The application provides a conversation method and device based on a syndication card, computer equipment and a computer readable storage medium. The method comprises the steps of obtaining voice input by a user, carrying out voice recognition on the voice to obtain a conversation text corresponding to the voice, carrying out natural language processing on the conversation text to obtain an entity and an intention contained in the conversation text, obtaining target display content corresponding to the intention according to the entity and the intention, calling a preset display control corresponding to the target display content according to the target display content, displaying the target display content in a card form through the preset display control, displaying the target display content in the card form, presenting the target display content in a structured form, and replacing content description of a text type with a rich text structured card to enable the content to be presented more clearly and easily, so that the situation that the simple text is not clear due to reply in the text form for user consultation is avoided, the user of the text with too long can not see the key point, the conversation efficiency can be improved, and according to the target display content, the preset display control corresponding to the target display content is called, the preset display control can be automatically assembled according to the target display content so as to combine different cards according to different target display contents to realize dynamic aggregation of the cards, namely, related contents can be assembled according to the information concerned by the user to flexibly construct the aggregate card, rather than writing the contents into the fixed card, thereby freely combining the preset display controls according to the actual target display content to be displayed, to support rich text complex typesetting, and can repeatedly call the same front-end display control to display different target display contents, any card can be obtained by using limited control combination, and convenience in display of the dialog text is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic flowchart of a conversation method based on syndication cards according to an embodiment of the present application;

fig. 2 is a schematic view of a first sub-flow of a conversation method based on a syndication card according to an embodiment of the present application

Fig. 3 is a schematic view of a second sub-flow of a conversation method based on syndication cards according to an embodiment of the present application;

fig. 4 is a schematic diagram of a syndication card in the syndication card based conversation method provided in the embodiment of the present application;

fig. 5 is a third sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application;

fig. 6 is a fourth sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application;

fig. 7 is a fifth sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application;

FIG. 8 is a schematic block diagram of a syndication card based dialog device as provided by an embodiment of the present application; and

fig. 9 is a schematic block diagram of a computer device provided in an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

Referring to fig. 1, fig. 1 is a schematic flowchart of a conversation method based on a syndication card according to an embodiment of the present application. As shown in FIG. 1, the method includes the following steps S11-S15:

and S11, acquiring the voice input by the user, and performing voice recognition on the voice to obtain a dialog text corresponding to the voice.

Specifically, in general, when a user performs self-service through a dialog system, a voice is input through a voice input device, and the voice input by the user needs to be subjected to voice recognition to convert the voice into a text, so as to obtain a dialog text corresponding to the voice input by the user.

And S12, performing natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text.

The Entity is also called Named Entity, and the english name is Named Entity, including name of person, organization name, place name and other entities identified by name, and the more extensive entities also include number, date, currency, address, etc.

Specifically, after a dialog text corresponding to self-service by a user is acquired, natural language processing including entity recognition and natural language understanding is performed on the dialog text, an entity included in the dialog text is extracted through the entity recognition, an intention corresponding to the dialog text is recognized through the natural language understanding, and then the user is fed back according to the intention, wherein the intention included in the dialog text is what the user intends to do in speech expression.

Further, entity recognition may be performed on the dialog text by using a preset entity recognition method to extract an entity included in the dialog text, an intention corresponding to a user for consultation is recognized, and then a reply is performed on the user according to the entity and the intention. The entity identification method comprises a traditional machine learning method, a deep learning method, an attention model, transfer learning, semi-supervised learning and the like. For example, when an entity is identified, a DSSM Model, that is, Deep Structured Semantic Model, may be used for entity linking, and a Deep neural network is used to represent text (sentences, Query, entities, etc.) into vectors, which are applied to a text similarity matching scenario.

Further, Natural Language Understanding (NLU for short) is to convert text information into semantic representation that can be processed by a machine, where the semantic representation usually includes first-order logic, semantic network, concept dependency, and frame-based representation, and semantic parsing is mainly performed through word segmentation, part-of-speech tagging, named entity recognition, syntactic analysis, and reference resolution to generate sentence meaning (i.e., to understand what the text is), perform intent recognition (typically through verb phrases, event mentions, such as querying weather), extract filling values of slots from the semantic representation, and then complete the semantic representation. Identifying user intent generally includes preprocessing error correction, word segmentation, syntactic analysis, entity linking, entity disambiguation, attribute or relationship identification. For example, the user asks "how many serious diseases the great fuxin 20 holds", the error correction module corrects the "great fuxin" to the correct name "great fuxing", and then the word is divided into { great fuxing 20, guarantee, how many, serious diseases }, wherein the great fuxing 20 has ambiguity and can represent the life insurance of the great fuxing 20 and the serious disease insurance of the great fuxing 20, but the life insurance of the great fuxing 20 is not connected with the serious disease, so that the user can be known to ask the entity of [ great fuxing 20 serious disease insurance ] in combination with the context. The statement of how many serious diseases are preserved can be identified by the model as the attribute of [ serious disease category ] of the query entity, so that the intention identification of the question of the user is completed by combining the identified entity and the attribute, and then the attribute of the entity can be inquired in the knowledge graph. Intent recognition may employ the TextCNN model.

Further, before the step of performing natural language processing on the dialog text to obtain the entity and the intention contained in the dialog text, the method further includes:

and preprocessing the dialog text.

Specifically, in order to improve the accuracy of recognizing the conversation intention and improve the quality and efficiency of replying to the user, after the text corresponding to the self-service of the user is acquired, preprocessing the dialog text, for example, preprocessing the character information contained in the dialog text such as correcting errors, removing stop words and segmenting words, and then performing entity recognition and natural language understanding on the preprocessed dialog text to recognize the entity contained in the dialog text and the intention corresponding to the dialog text, due to the fact that the text is obtained through preprocessing, errors possibly contained in the text are reduced, stop words are removed, interference of error information on entity recognition and natural language understanding is reduced, accuracy of recognition of intentions corresponding to the text is improved, and accuracy of replying the user conversation can be improved.

And S13, acquiring target display content corresponding to the intention according to the entity and the intention.

And S14, calling a preset display control corresponding to the target display content according to the target display content.

And S15, displaying the target display content in a card form through the preset display control.

Specifically, when a conversation between a person and a computer device is realized, for a situation corresponding to each problem that the person may present, the computer device prepares respective coping contents in advance for all situations, and each part of the coping contents contains a description of an entity. For example, for an insurance product a, the coping content corresponding to the insurance product a may be set, and the coping content of the insurance product a may include several parts such as an insurance product profile, insurance details, and application rules, and a corresponding content of a next level may be set for each part such as the insurance product profile, the insurance details, and the application rules.

After the entity and the intention are obtained, corresponding content corresponding to the entity is determined according to the entity, target display content corresponding to the intention is determined in the corresponding content corresponding to the entity, after the target display content corresponding to the intention is determined, a preset display control corresponding to the target display content is called according to the target display content, and the target display content is displayed in a card form through the preset display control. For example, if a user asks a question about what the application rule of the insurance product a is, the entity including the "insurance product a" and the "application rule" is obtained, and the intention is what the "application rule includes, the application rule may be determined according to the correspondence between the entity" insurance product a "and the insurance product a, then the specific content included in the application rule is used as the target display content, if the target display content includes the content C, a preset display control corresponding to the content C is called to display the content C, and the content C is displayed in the form of a card through the preset display control. If the target display content comprises three parts of EDF content, according to the three parts of EDF content included in the target display package, preset display controls corresponding to the EDF can be called according to preset, three preset cards are called to respectively display the three parts of EDF content, the three parts of EDF content are displayed in a card form through the corresponding preset display controls, and therefore according to the target display content, the preset display controls corresponding to the target display content are called, and the target display content is displayed in the card form through the preset display controls.

In the embodiment of the application, the voice input by a user is acquired, the voice is subjected to voice recognition to obtain a dialog text corresponding to the voice, the dialog text is subjected to natural language processing to obtain an entity and an intention contained in the dialog text, the target display content corresponding to the intention is acquired according to the entity and the intention, the preset display control corresponding to the target display content is called according to the target display content, the target display content is displayed in a card form through the preset display control, the target display content is displayed in the card form, the target display content can be presented in a structured form, the content description of a text type is replaced by a rich text structured card, the content presentation is clearer and easier, and the situation that the simple text is unclear due to the fact that the user consults and replies in the text form is avoided, the user of the text with too long can not see the key point, the conversation efficiency can be improved, and according to the target display content, the preset display control corresponding to the target display content is called, the preset display control can be automatically assembled according to the target display content so as to combine different cards according to different target display contents to realize dynamic aggregation of the cards, namely, related contents can be assembled according to the information concerned by the user to flexibly construct the aggregate card, rather than writing the contents into the fixed card, thereby freely combining the preset display controls according to the actual target display content to be displayed, to support rich text complex typesetting, and can repeatedly call the same front-end display control to display different target display contents, any card can be obtained by using limited control combination, and convenience of text display is improved.

Referring to fig. 2, fig. 2 is a schematic sub-flow chart of a conversation method based on syndication cards according to an embodiment of the present application. As shown in fig. 2, in this embodiment, the step of obtaining the target display content corresponding to the intention according to the entity and the intention includes:

s21, acquiring a preset knowledge graph corresponding to the entity according to the entity;

and S22, acquiring preset knowledge map triple contents corresponding to the intention from the preset knowledge map according to the intention, and taking the preset knowledge map triple contents as target display contents.

The Knowledge Graph, named Knowledge Graph in english, is a semantic network that exposes relationships between entities, and each piece of Knowledge in the Knowledge Graph may be described by an SPO triple, i.e., a Subject (i.e., Subject), a predicate (i.e., prediction) -Object (i.e., Object), for example, a triple may be described by (entity 1, relationship, entity 2) or (entity, attribute value).

Specifically, for each corresponding content, a knowledge graph corresponding to the corresponding content is preset, the corresponding content is described through the knowledge graph, an Entity (english is Entity) in a triplet of the knowledge graph can be used for describing an object contained in the corresponding content, a relationship (english is relationship) in the triplet is used for describing a relationship between the Entity and the Entity, each Entity corresponds to one card, the relationship in the triplet describes a structure between different cards, the Entity in the triplet can be used for describing the object contained in the corresponding content, an abstraction of the relationship between the Entity and the Entity is described through an attribute (english is Property) in the triplet, and a specific value corresponding to the attribute is described through an attribute value (PropertyValue) in the triplet. For example, for an insurance product a, a preset knowledge graph corresponding to the insurance product a may be set, where the insurance product a may include contents such as an insurance product profile, insurance details, and an insurance rule, where the insurance product a, the insurance product profile, the insurance details, and the insurance rule belong to entities, the insurance product a and the insurance product profile, the insurance details, and the insurance rule are logically linked by attribute description, and specific contents for each of the insurance product profile, the insurance details, and the insurance rule may be described by an attribute value, so as to construct the preset knowledge graph corresponding to the insurance product a.

And acquiring a preset knowledge graph triple content corresponding to the intention from the preset knowledge graph according to the entity after acquiring the preset knowledge graph corresponding to the entity, and taking the preset knowledge graph triple content as target display content, wherein the preset knowledge graph triple content comprises the relationship among different entities or comprises the entity, the attribute and the attribute value. For example, for an insurance product a, the entity included is identified as the "insurance product a" and the "guarantee details", the intention corresponding to the dialog text is identified as what the guarantee details of the insurance product a are, the entity included in the knowledge graph and the graph is located according to the entity "insurance product a" and the "guarantee details", and then the specific content corresponding to the "guarantee details" is located according to the corresponding intention, and the specific content corresponding to the "guarantee details" is the target display content.

Referring to fig. 3, fig. 3 is a second sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application. As shown in fig. 3, in this embodiment, the step of displaying the target display content in a card form through the preset display control includes:

s31, acquiring a target entity contained in the preset knowledge map triple content and an incidence relation corresponding to the target entity;

s32, filling the target entity into the preset display control for display;

s33, constructing a logic relation among different preset display controls through the association relation to form a card, and displaying the card.

Specifically, since the knowledge graph describes the relationship between entities, whether the knowledge graph is described in the form of a triple (entity 1, relationship, entity 2) or a triple (entity, attribute value), the essence of the knowledge graph is the description of the relationship between the entities, (entity 1, relationship, entity 2) can be understood as the relationship existing between parallel entities, and (entity, attribute value) can be understood as the relationship between inclusion relationship entities, that is, the relationship existing between entities, the attribute included in the entities, and the entity corresponding to the attribute value respectively exists as the inclusion relationship. Therefore, the preset knowledge graph triple content corresponding to the intention is obtained from the preset knowledge graph, so that the target entity contained in the preset knowledge graph triple content and the association relationship corresponding to the target entity can be identified, for example, for the insurance product a, the insurance product profile contained in the insurance product a and the specific content corresponding to the profile are the master-slave inclusion relationship, the insurance product a, the insurance product profile and the specific content corresponding to the profile are used as three target entities, and the relationship between the three target entities can be obtained.

The method comprises the steps of obtaining a target entity contained in preset knowledge graph triple content and an incidence relation corresponding to the target entity based on a knowledge graph, filling the target entity into a preset display control for display, using an automatic generation service to transcribe a formatted Json and send the Json to a client side as field values in the knowledge graph are stored in a database, filling the Json into the preset display control by the client side, so that a card structure is formed by reading graph relation edges and labels- > generating key-value pairs in the Json > filling and displaying the card by the client side according to the Json, and establishing a logical relation between different preset display controls through the incidence relation to form the card and display the card. For example, referring to fig. 4, fig. 4 is a schematic diagram of an aggregate card in the conversation method based on the aggregate card according to the embodiment of the present application, as shown in fig. 4, the card is constructed as an example, the text content in the diagram is entities, lines between the entities describe relationships between the entities, as can be seen from fig. 1, the product information (entities) includes a first-level classification 01 (entities) and a second-level classification 02 (entities), the first-level classification 01 (entities) includes the second-level classification 01-1 to the second-level classification 01-05, and so on.

In the embodiment of the application, the card structures based on different products are the same, for example, insurance products generally include modules such as profiles, guarantee details, insurance rules and the like, but the respective specific contents of different insurance products are different from product to product, so that the structured information is abstracted into a unified and reusable card structure, the structure can include a substructure so as to continue to expand, thereby realizing hierarchical structured information display, storing the hierarchical structure and the specific field value in a knowledge graph by means of the knowledge graph, using entities in the knowledge graph to correspond to nodes in the card structure, describing the hierarchical relationship between the nodes in the card structure through edge connection of triples of the knowledge graph, improving the efficiency and accuracy of dynamic aggregation of the card, and improving the construction effect of the card structure.

Referring to fig. 5, fig. 5 is a third sub-flowchart of a conversation method based on a syndication card according to an embodiment of the present application. As shown in fig. 5, in this embodiment, the target entity includes a hierarchical label corresponding to the target entity, and the step of obtaining the target entity included in the preset knowledge-graph triple content and the association relationship corresponding to the target entity includes:

s51, obtaining a hierarchical label corresponding to the target entity;

s52, establishing association relations among different target entities according to the corresponding hierarchical label sequence among the hierarchical labels.

The hierarchical label is an identifier describing the respective corresponding precedence relationship of different entities. Referring to fig. 1, in fig. 1, the hierarchical label corresponding to the product information is level 0, the hierarchical label corresponding to the product knowledge is level 01, the hierarchical label corresponding to the sales data is level 02, and so on.

Specifically, for a card structure including multiple levels, entities (i.e., field values) in a knowledge graph may be labeled with level labels of different levels to describe respective corresponding level relationships of precedence order between the entities, so as to establish a mapping relationship between the entities, and when the entities and the entity relationships in the knowledge graph are converted into nodes and relationships between the nodes included in the card, the relationships between the entities and the entities may be mapped by including Key-Value pairs in Json, where Key is a corresponding node name and Value is a group or a field Value. By obtaining the hierarchical labels corresponding to the target entities and establishing the incidence relation among different target entities according to the hierarchical label sequence corresponding to the hierarchical labels among the plurality of hierarchical labels, the incidence relation corresponding to the target entities contained in the preset knowledge map triple content is obtained, so that when the entities are used for constructing cards, the relation among the entities can be accurately identified.

In the embodiment of the present application, for a card structure including multiple levels, by presetting a level label corresponding to the target entity, several fixed controls may be used to form a card, including: the card controls comprise card titles, secondary titles, tables, pictures, item lists, icon groups, tertiary lists and the like, and the cards with rich text complex typesetting can be flexibly assembled and supported, so that any card can be obtained by using limited control combinations, please refer to fig. 1 continuously.

Referring to fig. 6, fig. 6 is a fourth sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application. As shown in fig. 6, in this embodiment, the target display content includes a main display content and a slave display content associated with the main display content, and the step of displaying the target display content in a card form through the preset display control includes:

s61, displaying the main display content through a first preset display control corresponding to the main display content by a card;

and S62, displaying the slave display content through a second preset display control corresponding to the slave display content by a card, and associating the second preset display control behind the corresponding first preset display control.

Specifically, in the embodiment of the present application, an intention corresponding to a user voice is used as a main display content, the main display content is displayed by a card through a first preset display control corresponding to the main display content, the main display content corresponds to a node to perform feedback on a user, therefore, the conversation is realized, meanwhile, in order to guide the user and improve the conversation efficiency, the slave display content related to the main display content is displayed in a card through the second preset display control corresponding to the slave display content, and after the second preset display control is associated with the corresponding first preset display control, the slave display content also corresponds to the node in the card, thereby realizing the recommendation of the associated node in the card, that is, when a user asks a question for a certain node in the structured information, the index of other nodes adjacent to the node and the root node can be prompted at the bottom after the question is solved. For example, the root node L has child nodes A, B, C, D and E, the user asks a question about C, and after a card about C is given, the neighbor node A, B, D, E of C is recalled in the background, and the root node L of C is ranked according to the historical click condition of the user, A, B, D, E is ranked so as to derive the content that the user is most interested in the current context, and the recall algorithm is based on the relationship between nodes in the graph (recall the neighbor node and the root node), wherein the recall is to select a part of the preset content as a candidate set.

Referring to fig. 7, fig. 7 is a fifth sub-flowchart of a conversation method based on syndication cards according to an embodiment of the present application. As shown in fig. 7, in this embodiment, the slave display content includes a plurality of sub-portion display contents, and the step of displaying the slave display content in a card through a second preset display control corresponding to the slave display content includes:

s71, acquiring a preset intimacy weight of each sub-part display content and the main display content;

s72, sequencing all the sub-part display contents according to the preset intimacy weights corresponding to the sub-part display contents from high to low to obtain a sequencing sequence corresponding to the sub-part display contents;

and S73, displaying all the sub-part display contents in a card through the corresponding preset display control according to the sorting sequence.

Specifically, after an entity contained in the dialog text is extracted, a preset main display content and a plurality of associated sub display contents which are matched with the entity and the intention are screened out from a preset knowledge graph according to the entity, the plurality of sub display contents are sequenced, and the preset main display content and the plurality of associated sub display contents are displayed to a user. For example, the root node L has child nodes A, B, C, D and E, the user asks a question for C, after a card for C is given, the neighboring node A, B, D, E of C is recalled in the background, and the root node L of C, for nodes C and A, B, D, E, preset affinity weights are respectively assigned to C, and according to the historical click condition of the user on each display content, the values of the affinity weights are learned and adjusted through an intelligent learning model, and the affinity weights are used for sequencing A, B, D, E, so that the content which is most concerned after the user knows C in the current context is deduced, the user is guided, and the conversation efficiency is improved. The historical click condition may include the historical click frequency of the user, the click frequencies of other users in the exhibition organization where the user is located, and the click frequencies of all users.

Further, in order to solve the data sparsity problem, the root nodes corresponding to the child nodes do not participate in the sorting, but the root nodes are fixed and pushed out, so that the user always has a path to return to the previous layer, and then clicks other concerned nodes from the previous layer.

In an embodiment, the entity includes a preset URL corresponding to a preset picture or a preset two-dimensional array corresponding to a preset table.

Specifically, the nodes in the card may be not only characters, but also contents such as pictures or tables. If the node in the card is a preset picture, the entity is a preset URL (Uniform Resource Locator) corresponding to the preset picture, if the node in the card is a preset table, the entity is a preset two-dimensional array corresponding to the preset table, when mapping from a knowledge graph structure to the card structure is realized through a Json file, the preset URL corresponding to the preset picture can be contained under a list corresponding to one node in the Json file, and the preset two-dimensional array representing the preset table is represented, so that the representation form of the card is enriched, the construction effect of the card structure is improved, the card is more visual, and the efficiency and communication quality of man-machine communication are improved.

In an embodiment, the entity further comprises a preset API interface.

Specifically, the preset API interface is used as an entity, so that other APIs are integrated into the card, for example, buttons can be integrated into the card, and other services can be called after clicking, so that rich functions and expression forms of the aggregated card are realized, the construction effect of the card structure is improved, the card is more visual, and the efficiency and the communication quality of man-machine communication are improved.

It should be noted that, the conversation method based on the aggregate card described in the above embodiments may recombine the technical features included in different embodiments as needed to obtain the combined implementation, but all of them are within the protection scope claimed in the present application.

Referring to fig. 8, fig. 8 is a schematic block diagram of a conversation device based on a syndication card according to an embodiment of the present application. Corresponding to the above conversation method based on the syndication card, the embodiment of the application also provides a conversation device based on the syndication card. As shown in fig. 8, the syndication card based dialog device includes means for performing the above-described syndication card based dialog method, which may be configured in a computer device. Specifically, referring to fig. 8, the conversation apparatus 80 based on the syndication card includes a first obtaining unit 81, a processing unit 82, a second obtaining unit 83, a retrieving unit 84 and a display unit 85.

The first obtaining unit 81 is configured to obtain a voice input by a user, and perform voice recognition on the voice to obtain a dialog text corresponding to the voice;

a processing unit 82, configured to perform natural language processing on the dialog text to obtain an entity and an intention included in the dialog text;

a second obtaining unit 83, configured to obtain, according to the entity and the intention, target display content corresponding to the intention;

the retrieving unit 84 is configured to retrieve a preset display control corresponding to the target display content according to the target display content;

and the display unit 85 is configured to display the target display content in a card form through the preset display control.

In an embodiment, the second obtaining unit 83 includes:

the first acquisition subunit is used for acquiring a preset knowledge graph corresponding to the entity according to the entity;

and the second obtaining subunit is configured to obtain, according to the intention, preset knowledge graph triple content corresponding to the intention from the preset knowledge graph, and use the preset knowledge graph triple content as target display content.

In one embodiment, the display unit 85 includes:

the third obtaining subunit is configured to obtain a target entity included in the preset knowledge graph triple content and an association relationship corresponding to the target entity;

the filling subunit is used for filling the target entity into the preset display control for display;

and the display subunit is used for constructing a logical relationship between different preset display controls through the association relationship to form a card and displaying the card.

In an embodiment, the target entity includes a hierarchical label corresponding to the target entity, and the third obtaining subunit includes:

a fourth obtaining subunit, configured to obtain a hierarchical label corresponding to the target entity;

and the establishing subunit is used for establishing an association relationship between different target entities according to the corresponding hierarchical label sequence among the hierarchical labels.

In one embodiment, the target display content includes a master display content and a slave display content associated with the master display content, and the display unit 85 includes:

the first display subunit is used for displaying the main display content in a card through a first preset display control corresponding to the main display content;

and the second display subunit is used for displaying the slave display content in a card through a second preset display control corresponding to the slave display content, and associating the second preset display control behind the corresponding first preset display control.

In one embodiment, the slave display content comprises a plurality of sub-portion display contents, and the second display sub-unit comprises:

a fifth acquiring subunit, configured to acquire a preset intimacy degree weight between each of the sub-portion display contents and the main display content;

the sorting subunit is configured to sort all the sub-portion display contents according to the preset affinity weights corresponding to the sub-portion display contents in the order from high to low, so as to obtain a sorting sequence corresponding to the sub-portion display contents;

and the third display subunit is used for displaying all the sub-part display contents in a card through the corresponding preset display control according to the sorting sequence.

In an embodiment, the entity includes a preset URL corresponding to a preset picture, a preset two-dimensional array corresponding to a preset table, and a preset API interface.

It should be noted that, as can be clearly understood by those skilled in the art, the specific implementation process of the above dialog device and each unit based on the aggregated card may refer to the corresponding description in the foregoing method embodiment, and for convenience and brevity of description, no further description is provided herein.

Meanwhile, the division and connection manner of each unit in the conversation device based on the aggregate card are only used for illustration, in other embodiments, the conversation device based on the aggregate card may be divided into different units as required, or each unit in the conversation device based on the aggregate card may adopt different connection sequences and manners, so as to complete all or part of the functions of the conversation device based on the aggregate card.

The above-described conversation apparatus based on the aggregate card may be implemented in the form of a computer program that can be run on a computer device as shown in fig. 9.

Referring to fig. 9, fig. 9 is a schematic block diagram of a computer device according to an embodiment of the present application. The computer device 500 may be a computer device such as a desktop computer or a server, or may be a component or part of another device.

Referring to fig. 9, the computer device 500 includes a processor 502, a memory, which may include a non-volatile storage medium 503 and an internal memory 504, which may also be a volatile storage medium, and a network interface 505 connected by a system bus 501.

The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, causes the processor 502 to perform one of the above-described syndication card-based conversation methods.

The processor 502 is used to provide computing and control capabilities to support the operation of the overall computer device 500.

The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503, and when the computer program 5032 is executed by the processor 502, the processor 502 can execute a method for conversation based on the aggregate card.

The network interface 505 is used for network communication with other devices. Those skilled in the art will appreciate that the configuration shown in fig. 9 is a block diagram of only a portion of the configuration associated with the present application and does not constitute a limitation of the computer device 500 to which the present application may be applied, and that a particular computer device 500 may include more or less components than those shown, or may combine certain components, or have a different arrangement of components. For example, in some embodiments, the computer device may only include a memory and a processor, and in such embodiments, the structures and functions of the memory and the processor are consistent with those of the embodiment shown in fig. 9, and are not described herein again.

Wherein the processor 502 is configured to run the computer program 5032 stored in the memory to implement the following steps: acquiring voice input by a user, and performing voice recognition on the voice to obtain a dialog text corresponding to the voice; carrying out natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text; according to the entity and the intention, target display content corresponding to the intention is obtained; calling a preset display control corresponding to the target display content according to the target display content; and displaying the target display content in a card form through the preset display control.

In an embodiment, when the processor 502 implements the step of obtaining the target display content corresponding to the intention according to the entity and the intention, the following steps are specifically implemented:

acquiring a preset knowledge graph corresponding to the entity according to the entity;

and acquiring preset knowledge map triple contents corresponding to the intention from the preset knowledge map according to the intention, and taking the preset knowledge map triple contents as target display contents.

In an embodiment, when the processor 502 implements the step of displaying the target display content in the form of a card through the preset display control, the following steps are specifically implemented:

acquiring a target entity contained in the preset knowledge map triple content and an incidence relation corresponding to the target entity;

filling the target entity into the preset display control for display;

and constructing a logic relationship between different preset display controls through the association relationship to form a card, and displaying the card.

In an embodiment, the target entity includes a hierarchical label corresponding to the target entity, and when the step of obtaining the target entity included in the preset knowledge graph triple content and the association relationship corresponding to the target entity is implemented, the processor 502 specifically implements the following steps:

acquiring a hierarchical label corresponding to the target entity;

and establishing an incidence relation between different target entities according to the corresponding hierarchical label sequence among the hierarchical labels.

In an embodiment, the target display content includes a main display content and a slave display content associated with the main display content, and when the processor 502 implements the step of displaying the target display content in a card form through the preset display control, the following steps are specifically implemented:

displaying the main display content by a card through a first preset display control corresponding to the main display content;

and displaying the slave display content in a card through a second preset display control corresponding to the slave display content, and associating the second preset display control behind the corresponding first preset display control.

In an embodiment, when the processor 502 implements the step that the slave display content includes a plurality of sub-portion display contents, and the slave display content is displayed in a card through a second preset display control corresponding to the slave display content, the following steps are specifically implemented:

acquiring a preset intimacy weight of each sub-part display content and the main display content;

sequencing all the sub-part display contents according to the preset intimacy weights corresponding to the sub-part display contents from high to low to obtain a sequencing sequence corresponding to the sub-part display contents;

and displaying all the sub-part display contents in a card through the corresponding preset display control according to the sequencing sequence.

It should be understood that in the embodiment of the present Application, the Processor 502 may be a Central Processing Unit (CPU), and the Processor 502 may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

It will be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiments may be implemented by a computer program, and the computer program may be stored in a computer readable storage medium. The computer program is executed by at least one processor in the computer system to implement the flow steps of the embodiments of the method described above.

Accordingly, the present application also provides a computer-readable storage medium. The computer-readable storage medium may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, the computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of:

a computer program product which, when run on a computer, causes the computer to perform the steps of the syndication card based dialog method described in the embodiments above.

The computer readable storage medium may be an internal storage unit of the aforementioned device, such as a hard disk or a memory of the device. The computer readable storage medium may also be an external storage device of the device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the device. Further, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the apparatus.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The storage medium is an entity and non-transitory storage medium, and may be various entity storage media capable of storing computer programs, such as a usb disk, a removable hard disk, a Read-only memory (ROM), a magnetic disk, or an optical disk.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative. For example, the division of each unit is only one logic function division, and there may be another division manner in actual implementation. For example, various elements or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented.

The steps in the method of the embodiment of the application can be sequentially adjusted, combined and deleted according to actual needs. The units in the device of the embodiment of the application can be combined, divided and deleted according to actual needs. In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or contributed to by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing an electronic device (which may be a personal computer, a terminal, or a network device) to perform all or part of the steps of the method according to the embodiments of the present application.

The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, and these modifications or substitutions should be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A syndication card-based conversation method, the method comprising:

acquiring voice input by a user, and performing voice recognition on the voice to obtain a dialog text corresponding to the voice;

carrying out natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text;

according to the entity and the intention, target display content corresponding to the intention is obtained;

calling a preset display control corresponding to the target display content according to the target display content;

and displaying the target display content in a card form through the preset display control.

2. The syndication card-based conversation method according to claim 1, wherein the step of obtaining the target display content corresponding to the intention according to the entity and the intention comprises:

3. The aggregated card-based dialog method of claim 1, wherein the step of displaying the target display content in card form through the preset display control comprises:

filling the target entity into the preset display control for display;

4. The aggregated card based dialog method according to claim 3, wherein the target entity includes a hierarchical label corresponding to the target entity, and the step of obtaining the target entity included in the preset knowledge-graph triple content and the association relationship corresponding to the target entity includes:

acquiring a hierarchical label corresponding to the target entity;

5. The aggregated card-based dialog method according to claim 1, wherein the target display content comprises a master display content and a slave display content associated with the master display content, and the step of displaying the target display content in a card form through the preset display control comprises:

6. The syndication card-based conversation method according to claim 5, wherein the slave display content includes a plurality of sub-portion display content, and the step of displaying the slave display content as a card through a second preset display control corresponding to the slave display content comprises:

7. The card aggregation-based dialog method of claim 1, wherein the entity comprises a preset URL corresponding to a preset picture, a preset two-dimensional array corresponding to a preset table, and a preset API interface.

8. A syndicated card-based conversation device, comprising:

the device comprises a first acquisition unit, a second acquisition unit and a processing unit, wherein the first acquisition unit is used for acquiring voice input by a user and carrying out voice recognition on the voice to obtain a dialog text corresponding to the voice;

the processing unit is used for carrying out natural language processing on the dialog text to obtain an entity and an intention contained in the dialog text;

the second acquisition unit is used for acquiring target display content corresponding to the intention according to the entity and the intention;

the calling unit is used for calling a preset display control corresponding to the target display content according to the target display content;

and the display unit is used for displaying the target display content in a card form through the preset display control.

9. A computer device, comprising a memory and a processor coupled to the memory; the memory is used for storing a computer program; the processor is adapted to run the computer program to perform the steps of the method according to any of claims 1-7.

10. A computer-readable storage medium, characterized in that the storage medium stores a computer program which, when being executed by a processor, realizes the steps of the method according to any one of claims 1 to 7.