Seat service response processing method and device and related equipment

Abstract

The invention belongs to the technical field of financial science and technology, provides a seat service response processing method and device, computer equipment and a computer readable storage medium, and aims to solve the problem of low online service efficiency of a customer service system of financial science and technology in the traditional technology. Based on a preset knowledge graph, automatically answering the question to obtain interactive content, matching the user content with the content in the preset knowledge graph to obtain the similarity between the user content and the corresponding content in the preset knowledge graph, and performing priority labeling on the call according to the similarity to obtain the user content and the corresponding content in the preset knowledge graph. And responding to an instruction of acquiring the call by the seat, pushing the call to the seat according to a sequence of the priorities from high to low, and making a manual response to the call by the seat, so that the customer list contact efficiency of seat operation can be improved, and the call service efficiency of a customer service system is improved.

Description

Technical Field

[0001] The present application relates to the field of financial technology, and in particular to a method, device, computer equipment and computer-readable storage medium for agent service response processing. Background Art

[0002] In the field of financial technology, for example, insurance companies have many insurance products such as auto insurance and non-auto insurance, and there are also a large number of users. In order to provide online services to users, an online customer service system is generally used. The online customer service system can be referred to as a customer service system, and seats are configured to provide online services to users based on the customer service system and through seats.

[0003] Generally speaking, since there are many users who consult online, and the number of seats is limited, the customer service system will allocate a certain number of customer lists to each seat. However, the number of seats is not proportional to the number of customer lists (i.e., users). The more customers there are, the more customer lists each seat will be allocated when the number of seats is fixed. The seats provide services to users based on the corresponding number of customer lists directly allocated by the customer service system. Therefore, the seats have many customer lists to contact every day to perform customer service operations.

[0004] When performing customer service operations, if the agents contact the customer lists directly assigned by the customer service system, especially when the number of customer lists is relatively large, it is time-consuming and labor-intensive for the agents to do so. Even when the number of customer lists reaches a certain level, the agents may not be able to contact all the customer lists, resulting in low efficiency in the agents contacting the customer lists, reducing the customer service efficiency of the agents' operations, and reducing the online service efficiency of the customer service system. Summary of the invention

[0005] The present application provides a seat service response processing method, device, computer equipment and computer-readable storage medium, which can solve the technical problem of low online service efficiency of customer service systems in traditional technologies.

[0006] In a first aspect, the present application provides a seat service answer processing method, including: responding to a call initiated by a user online, obtaining a question contained in the call; automatically answering the question based on a preset knowledge graph to obtain interactive content, wherein the interactive content includes user content sent by a user side; matching the user content with the content in the preset knowledge graph to obtain a similarity between the user content and the corresponding content in the preset knowledge graph; marking the call with a priority based on the similarity to obtain the priority corresponding to the call; responding to the seat's instruction to obtain the call, pushing the call to the seat in order of priority from high to low, and then the seat manually answers the call.

[0007] In a second aspect, the present application provides a seat service answer processing device, including: a first acquisition unit, used to respond to a call initiated by a user online, and obtain a question contained in the call; an automatic answer unit, used to automatically answer the question based on a preset knowledge graph, and obtain interactive content, wherein the interactive content includes user content sent by the user side; a first matching unit, used to match the user content with the content in the preset knowledge graph, and obtain the similarity between the user content and the corresponding content in the preset knowledge graph; a labeling unit, used to label the call with a priority according to the similarity, and obtain the priority corresponding to the call; an agent answering unit, used to respond to the agent's instruction to obtain the call, and push the call to the agent in order of priority from high to low, so that the agent manually answers the call.

[0008] In a third aspect, the present application provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the agent service response processing method when executing the computer program.

[0009] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the agent service response processing method.

[0010] The present application provides a seat service answer processing method, device, computer equipment and computer-readable storage medium. The method responds to a call initiated by a user online, obtains a question contained in the call, automatically answers the question based on a preset knowledge graph, obtains interactive content, and the interactive content includes user content sent by the user side, matches the user content with the content in the preset knowledge graph, obtains the similarity between the user content and the corresponding content in the preset knowledge graph, and marks the call with priority according to the similarity to obtain the priority corresponding to the call, responds to the seat's instruction to obtain the call, pushes the call to the seat in order from high to low priority, and then the agent manually answers the call. By matching the user content in the automatic answering process with the content in the preset knowledge graph, the real intention of the online user can be screened to give priority to serving high-quality users whose real intention of the user matches the actual needs of the business. Compared with the traditional technology of randomly pushing online users to the seat, the customer list contact efficiency of the seat operation can be improved, thereby improving the call service efficiency of the customer service system and creating greater customer service value. For example, when applied in the customer service scenario of an insurance company providing insurance products, the call service efficiency of the seat service of the insurance product can be improved, and greater insurance customer service value can be created. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0012] Figure 1 A flowchart of a seat service response processing method provided in an embodiment of the present application;

[0013] Figure 2 A schematic diagram of the first sub-flow of the agent service response processing method provided in an embodiment of the present application;

[0014] Figure 3 A schematic diagram of a second sub-flow of the agent service response processing method provided in an embodiment of the present application;

[0015] Figure 4 A schematic block diagram of an agent service response processing device provided in an embodiment of the present application;

[0016] Figure 5 A schematic block diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0019] An embodiment of the present application provides a method for processing agent service responses, which can be applied to but not limited to computer devices such as laptops, desktop computers, and servers, and is used in the field of financial technology, including but not limited to, when agents provide online services to users based on a customer service system to perform response processing. For example, it is used in customer service scenarios where insurance companies provide insurance products to provide online customer service for insurance users.

[0020] In the face of the technical problem of low online service efficiency of customer service systems in traditional technologies, the inventors proposed an agent service response processing method of an embodiment of the present application. The core idea of ​​the embodiment of the present application is: pre-construct a knowledge graph, automatically respond to users based on the knowledge graph, and classify users according to the interaction between the automatic response and the user, so as to automatically screen out the user's true intentions. When the agent provides services to online users, the user is pushed to the agent according to the priority of the user classification, and the agent provides online services to the pushed users, thereby filtering out invalid lists or low-quality lists in online calls according to the automatic response and user classification, and providing online services to high-priority users first. Compared with the traditional technology of randomly pushing online users to agents, the customer list contact efficiency of the agent's operation can be improved, thereby improving the call service efficiency of the customer service system and creating greater service value.

[0021] Some embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0022] See also Figure 1 , Figure 1 The following is a flow chart of the seat service response processing method provided in the embodiment of the present application. Figure 1 As shown, in this embodiment, the method includes but is not limited to the following steps S11-S15:

[0023] S11. Respond to a call initiated by an online user and obtain a question sentence included in the call.

[0024] Explanatory, in online service scenarios, users generally initiate online calls only when they have questions to consult. When users initiate online calls, they generally ask questions directly, and the questions are generally expressed in the form of questions. For example, in the field of financial technology, in the online service scenarios of insurance companies for insurance products, users generally initiate online calls based on their needs, and generally users initiate calls with questions of a consulting nature. In this case, the questions contained in the call are obtained. For example, for insurance product A, users may ask "Can my occupation be insured with insurance product A?", "What are the main coverage contents of product insurance A?", etc.

[0025] S12. Based on a preset knowledge graph, the question is automatically answered to obtain interactive content, where the interactive content includes user content sent by the user side.

[0026] Explanatory, pre-constructing knowledge graphs, i.e., preset knowledge graphs. For example, in the field of financial technology, in the online service scenarios of insurance companies for insurance products, insurance product knowledge can be extracted from unstructured text represented by triples (entity 1-relationship-entity 2) based on Bi-LSTM-CRF joint automatic extraction. Insurance product knowledge includes but is not limited to the insurance company name, insurance product name, insurance product link, claim information, insurance age, guarantee period, main protection, underwriting, renewal, claim settlement, etc. extracted from insurance terms, and then the insurance product knowledge extracted from multiple sources and heterogeneity is integrated to ensure the completeness and consistency of insurance product knowledge. Finally, the integrated knowledge is stored in the form of a knowledge graph to obtain a preset knowledge graph for insurance products.

[0027] When providing online services and automatically answering questions based on a preset knowledge graph, the questions contained in the call are obtained and matched with the preset knowledge graph. When the questions match the preset knowledge graph, the preset answers in the preset knowledge graph are used to respond to the user's questions for intelligent automatic answering. Such automatic answering generally has n round trips, where n≥1, and n is a natural number, to enable the automatic answering to interact with the user and obtain interactive content. The interactive content includes the content sent by the user on the user side, that is, the user content.

[0028] S13. Match the user content with the content in the preset knowledge graph to obtain the similarity between the user content and the corresponding content in the preset knowledge graph.

[0029] Explanatory, when constructing a preset knowledge graph, a user-side content is predicted and constructed in advance according to the call scenarios in actual business, and the user-side content includes but is not limited to the questions asked by the user. Exemplarily, in the field of financial technology, in the online service scenario of insurance product A by an insurance company, when constructing a preset knowledge graph of insurance product A, a user-side content is predicted and constructed in advance according to the call scenarios of insurance product A in actual business, that is, what content the user will send about insurance product A, and the user-side content includes but is not limited to the questions sent by the user about insurance product A, and also includes the company name, product name, underwriting age, protection period, insured age, underwriting, renewal and other descriptions of insurance product A sent by the user, so as to predict and collect as much content as possible sent by the user on the topic of insurance A, and obtain the content that the user may send about insurance product A in the preset knowledge graph of insurance product A.

[0030] Based on the construction of the above-mentioned preset knowledge graph, the user content sent by the user side during the automatic answering process is matched with the relevant content in the preset knowledge graph, that is, matched with the content predicted in the preset knowledge graph that the user may send. The two can be matched at the entity and / or sentence levels respectively. The similarity calculation can use methods including but not limited to Euclidean distance, Manhattan distance, Chebyshev distance, angle cosine, and K-Means clustering algorithm to obtain the similarity between the user content and the corresponding content in the preset knowledge graph.

[0031] S14. Prioritize the calls according to the similarity to obtain the priorities corresponding to the calls.

[0032] Explanatory, based on the similarity, the calls are prioritized to obtain the corresponding priorities of the calls. The priorities can be represented by labels such as scores, categories or levels, where the similarity is positively correlated with the priority, that is, the higher the similarity, the higher the priority, indicating that the content of the user's call is closer to the content that the user may send predicted in the preset knowledge graph, and the user's true intention is closer to the actual needs of the business; the lower the similarity, the lower the priority, indicating that the content of the user's call deviates more from the content that the user may send predicted in the preset knowledge graph, and the user's true intention deviates more from the actual needs of the business.

[0033] S15. In response to the agent's instruction to obtain a call, the call is pushed to the agent in order of priority from high to low, and the agent manually answers the call.

[0034] Explanatoryly, the front-end agent performs online service. After serving an online user, the front-end agent obtains the next online call through the user acquisition button on the system interface, responds to the agent's instruction to obtain the call, and pushes the call to the agent in order of priority from high to low, and the agent then manually answers the call.

[0035] In an embodiment of the present application, by responding to a call initiated by a user online, a question included in the call is obtained, and based on a preset knowledge graph, the question is automatically answered to obtain interactive content, the interactive content includes user content sent by the user side, the user content is matched with the content in the preset knowledge graph, and the similarity between the user content and the corresponding content in the preset knowledge graph is obtained. According to the similarity, the call is prioritized to obtain the priority corresponding to the call, and the agent is responded to to obtain the call instruction, and the call is pushed to the agent in order of priority from high to low, and then the agent manually answers the call. By matching the user content in the automatic answering process with the content in the preset knowledge graph, the real intention of the online user can be screened to give priority to serving high-quality users whose real intention of the user matches the actual needs of the business. Compared with the traditional technology of randomly pushing online users to the agent, the customer list contact efficiency of the agent operation can be improved, thereby improving the call service efficiency of the customer service system and creating greater service value. For example, when applied in a customer service scenario where an insurance company provides insurance products, the call service efficiency of the agent service of the insurance product can be improved, and greater insurance customer service value can be created.

[0036] In one embodiment, see Figure 2 , Figure 2 This is a schematic diagram of the first sub-flow of the seat service response processing method provided in the embodiment of the present application. Figure 2 As shown, in this embodiment, based on the preset knowledge graph, the question is automatically answered to obtain interactive content, including:

[0037] S21, identifying entities contained in the question sentence to obtain user entities;

[0038] S22, matching the user entity with entities included in a preset knowledge graph;

[0039] S23. When there is a match between the entity contained in the preset knowledge graph and the user entity, obtain the graph entity corresponding to the user entity in the preset knowledge graph, and link the user entity to the graph entity;

[0040] S24, matching the question with a preset question associated with the graph entity based on the link between the user entity and the graph entity;

[0041] S25. When the question sentence matches the preset question sentence, a preset answer corresponding to the preset question sentence is provided to the user to obtain interactive content.

[0042] Explanatory, entities, also known as named entities, generally refer to objects with specific meanings or strong referentiality in texts, usually including names of people, places, institutions, dates, times, proper nouns, etc. Entity recognition, also known as named entity recognition (NER for short), also known as "proper name recognition", usually includes two parts: entity boundary recognition; determining the type of entity (name, place name, institution name or other).

[0043] To identify the entities contained in the questions sent by the user, a trained algorithm model can generally be used for identification. For example, a trained Bi-LSTM-CRF model is used to identify the entities in the questions sent by the user to obtain user entities. The user entity describes the entity on the user side, or is called the entity contained in the content sent by the user. It should be noted that in the embodiment of the present application, "user" is used to describe the entity, not to limit the entity, but to distinguish different entities, and other similar situations are similar.

[0044] The entities in the questions asked by the user are identified. After obtaining the user entity, the user entity is matched with the entities contained in the preset knowledge graph. Generally, the similarity between the user entity and the entities contained in the preset knowledge graph is calculated to determine whether the user entity matches the entities contained in the preset knowledge graph. When the similarity is greater than or equal to the preset similarity threshold, it is determined that the user entity matches the entity contained in the preset knowledge graph. When the similarity is less than the preset similarity threshold, it is determined that the user entity does not match the entity contained in the preset knowledge graph.

[0045] When there is a match between the entities included in the preset knowledge graph and the user entity, the graph entity corresponding to the user entity in the preset knowledge graph is obtained, and the user entity is linked to the graph entity, wherein the graph entity describes the entity included in the preset knowledge graph. It should be noted that in the embodiment of the present application, the term "graph" is used to describe the entity, not to limit the entity, but to distinguish different entities, and other similar situations are similar.

[0046] Based on the link between the user entity and the graph entity, after matching at the entity level, the question is matched with the preset question associated with the graph entity. It should be noted that when constructing a preset knowledge graph, questions associated with the graph entity, namely preset questions, are pre-set. The preset questions describe several questions that users generally ask for the graph entity, and these questions can be obtained from the customer service's historical service data.

[0047] When a question matches a preset question, the preset answer corresponding to the preset question is responded to the user to obtain interactive content. It should be noted that when constructing a preset knowledge graph, the answer associated with the preset question, that is, the preset answer, is preset. The preset answer describes how to answer the question corresponding to the graph entity. Therefore, the preset knowledge graph at least includes a knowledge chain of "graph entity-preset question-preset answer", and the correspondence of "graph entity-preset question-preset answer" can be one-to-one, one-to-many, many-to-many, or many-to-one according to the specific business, that is, a single graph entity can correspond to several preset questions, a single preset question can correspond to several preset answers, a single preset answer can correspond to several preset questions, and a single preset question can correspond to several graph entities.

[0048] In an embodiment of the present application, the user entity is obtained by identifying the entity contained in the question, and the user entity is matched with the entity contained in the preset knowledge graph. When the entity contained in the preset knowledge graph matches the user entity, the graph entity corresponding to the user entity in the preset knowledge graph is obtained, and then the user entity is linked to the graph entity. Based on the link between the user entity and the graph entity, the question sentence is matched with the preset question sentence associated with the graph entity, and when the question sentence matches the preset question sentence, the preset answer corresponding to the preset question sentence is responded to the user to obtain interactive content, thereby matching the question sent by the user with the preset knowledge graph at two levels, namely, the entity and the question sentence. Since the complexity of matching entities and questions is different, not only the efficiency of matching the question sent by the user with the preset knowledge graph can be improved, but also the accuracy of matching the question sent by the user with the preset knowledge graph can be improved, thereby improving the efficiency and accuracy of automatic answering, improving the quality of automatic answering, and improving the call efficiency and quality of agent service.

[0049] In one embodiment, see Figure 3 , Figure 3 This is a schematic diagram of the second sub-flow of the seat service response processing method provided in the embodiment of the present application. Figure 3 As shown, in this embodiment, according to the similarity, the call is marked with a priority, and the priority corresponding to the call is obtained, including:

[0050] S31, determining whether the similarity is less than or equal to a preset similarity threshold;

[0051] S32. When the similarity is greater than a preset similarity threshold, the call is marked with a priority to obtain a priority corresponding to the call;

[0052] S33: When the similarity is less than or equal to a preset similarity threshold, determine that the call is an invalid call.

[0053] Explanatory, it is determined whether the similarity is less than or equal to the preset similarity threshold. When the similarity is greater than the preset similarity threshold, it indicates that the content of the user's call is closer to the content that the user may send predicted in the preset knowledge graph, and the user's real intention is closer to the actual needs of the business. The higher the similarity, the closer the above two are, until the above two are exactly the same. The higher the similarity, the call is marked with high priority to obtain the priority corresponding to the call. When the similarity is less than or equal to the preset similarity threshold, it indicates that the content of the user's call deviates more from the content that the user may send predicted in the preset knowledge graph, that is, the content of the user's call is more different from the content that the user may send predicted in the preset knowledge graph, and the user's real intention deviates more from the actual needs of the business. The lower the similarity, the greater the degree of deviation between the above two, until the above two are completely different. The lower the similarity, the call is marked with low priority, and the call is determined to be an invalid call or a low-quality call. The invalid call or the low-quality call is processed later or not processed, thereby filtering out the invalid call list or the low-quality call list in the online call, thereby improving the time efficiency ratio of the seat manual service, and then improving the call service efficiency of the customer service system, creating greater service value.

[0054] In the embodiment of the present application, by judging whether the similarity is less than or equal to a preset similarity threshold, when the similarity is greater than the preset similarity threshold, the call is marked with a priority to obtain the priority corresponding to the call, and when the similarity is less than or equal to the preset similarity threshold, the call is determined to be an invalid call, thereby prioritizing online users according to the content of the automatic answer in the online service to filter out the invalid call list or the low-quality call list in the online call. Compared with the traditional technology of treating online users equally, the customer list contact efficiency of subsequent agent operations can be improved, thereby improving the call service efficiency of the customer service system.

[0055] Further, after determining that the call is an invalid call, the method further includes:

[0056] The user is prompted to change the way of asking questions, and after being prompted a preset number of times, the call is terminated.

[0057] Explanatoryly, due to the complexity and diversity of user language expressions and the limitations of the algorithm in judging calls, after a call is determined to be an invalid call, the user is further prompted to change the way of asking questions, that is, the way of expression, form, and content of the question, so that the user can change the question for consultation and communication. At the same time, in order to avoid infinitely prompting the user to change the question, the call is interrupted after a preset number of prompts, thereby avoiding too much invalid communication with the same user.

[0058] Furthermore, before the call is interrupted, the user may be prompted to consult or communicate in other ways including but not limited to telephone and offline on-site.

[0059] The embodiment of the present application prompts the user to change the way of asking questions and interrupts the call after being prompted a preset number of times. This can improve the quality of automatic answering by changing the way of asking questions a limited number of times, thereby avoiding misjudgment of the call as an invalid call, and further improve the call service efficiency of the customer service system.

[0060] In one embodiment, obtaining the question sentence included in the call includes:

[0061] Determining whether the content sent online by the user is a question;

[0062] When the content is not a question, the user is prompted to send a question for consultation.

[0063] Explanation: Since users usually initiate online consultations because they have questions or problems that need to be clarified or verified, users usually come to ask the service staff to answer their questions. However, due to the complexity and diversity of user language expression, even if users come to the call to get answers to their questions, when users express their questions, they may not necessarily express them in the form of questions, but may also use statements to describe the existing problems.

[0064] For example, in insurance products in the field of financial technology, for a certain insurance product, a user's usual language expression for the question of uninsurable occupations may be "I am in occupation A, can I be insured by insurance product B?", but some users may express the same question as "I am in occupation A, and I'm not sure whether I can be insured by insurance product B". Taking this as an example, although the users express the same intention, their language expressions are different. The former is in the form of a question, and the latter is in the form of a statement. Obviously, the question can more clearly reflect the user's true intention, and the expression form of the statement, although it is not a problem for people to understand, is less certain for the algorithm model.

[0065] When determining the user's intention through an algorithm model, the more certain the expression is, the more conducive it is to the recognition accuracy of the algorithm model. Therefore, in order to allow the algorithm model to more accurately understand the user's true intention, the user is encouraged to express it in a manner that conforms to the specifications or standards of algorithm pattern recognition. Therefore, in an embodiment of the present application, it is determined whether the content sent by the user online is a question. If the content is not a question, the user is prompted to send a consultation question. If the content sent by the user is a question, the content is processed directly according to the user's question.

[0066] The embodiment of the present application determines whether the content sent by the user online is a question. If the content is not a question, the user is prompted to send a consultation question. This can further improve the accuracy of understanding the user's intention, improve the quality of automatic answering, and further improve the call service efficiency of the customer service system.

[0067] Furthermore, when the content is not a question, prompting the user to send a consultation question includes:

[0068] In the case where the content is not a question sentence, determining whether the content is a preset honorific;

[0069] In the case that the content is not a preset honorific, the user is prompted to send a consultation question.

[0070] Explanation: honorifics refer to the language means of showing respect to the listener. In different language environments, the use of honorifics will reflect different etiquette. Honorifics include: (1) greeting each other. When people meet each other, they greet each other with "Hello!" or "Good morning." Here, one word conveys at least three meanings to the other party: showing respect, showing kindness, and giving friendship. At the same time, it also shows three characteristics of oneself: being well-educated, graceful, and polite; (2) thanking others for their hard work. After the other party gives help, support, care, respect, and praise, the most concise, timely, and effective response is to say "thank you" from the bottom of your heart.

[0071] During the online service process, users are not sure whether the other party is a human or an intelligent robot, or out of language habits, they often use honorifics to start or end online communication. After the user sends the content of the communication, if the content is not a question, determine whether the content is a preset honorific. For example, the preset honorific includes but is not limited to "Hello", "Thank you", "Thank you very much", and "Goodbye". If the content is not a preset honorific, prompt the user to send the consultation question. If the content is a preset honorific, it should generally be responded with honorifics, indicating that the user has started the call or is about to end the call, and there is no need to prompt the user to send the consultation question.

[0072] In the embodiment of the present application, by determining whether the content is a preset honorific when the content is not a question, and prompting the user to send a consultation question when the content is not a preset honorific, the accuracy of understanding the user's intention can be further improved, thereby improving the quality of automatic answering and further improving the call service efficiency of the customer service system.

[0073] It should be noted that the agent service response processing method described in the above embodiments can recombine the technical features contained in different embodiments as needed to obtain a combined implementation plan, but all of them are within the protection scope required by this application.

[0074] See also Figure 4 , Figure 4 This is a schematic block diagram of a seat service response processing device provided in an embodiment of the present application. Corresponding to the above-mentioned seat service response processing method, an embodiment of the present application also provides a seat service response processing device. Figure 4 As shown, the seat service response processing device includes a unit for executing the above-mentioned seat service response processing method, and the seat service response processing device can be configured in a computer device. Figure 4 The seat service response processing device 40 includes a first acquisition unit 41, an automatic response unit 42, a first matching unit 43, a marking unit 44 and a seat response unit 45.

[0075] The first acquisition unit 41 is used to respond to a call initiated by an online user and acquire a question sentence included in the call;

[0076] An automatic answering unit 42, configured to automatically answer the question based on a preset knowledge graph to obtain interactive content, wherein the interactive content includes user content sent by the user side;

[0077] A first matching unit 43, configured to match the user content with the content in the preset knowledge graph to obtain a similarity between the user content and the corresponding content in the preset knowledge graph;

[0078] A labeling unit 44, configured to label the call with a priority according to the similarity, so as to obtain a priority corresponding to the call;

[0079] The agent answering unit 45 is used to respond to the agent's instruction to obtain a call, and push the call to the agent in descending order of priority, so that the agent manually answers the call.

[0080] In one embodiment, the automatic answering unit 42 includes:

[0081] A first recognition subunit is used to recognize entities contained in the question sentence to obtain user entities;

[0082] A first matching subunit, used to match the user entity with entities included in a preset knowledge graph;

[0083] A linking subunit, used to obtain the graph entity corresponding to the user entity in the preset knowledge graph and link the user entity to the graph entity when there is a match between the entity contained in the preset knowledge graph and the user entity;

[0084] A second matching subunit, configured to match the question with a preset question associated with the graph entity based on a link between the user entity and the graph entity;

[0085] The automatic answering subunit is used to answer the user with a preset answer corresponding to the preset question when the question matches the preset question, so as to obtain interactive content.

[0086] In one embodiment, the labeling unit 44 includes:

[0087] A first judging subunit, used to judge whether the similarity is less than or equal to a preset similarity threshold;

[0088] The marking subunit is used to mark the priority of the call when the similarity is greater than a preset similarity threshold, so as to obtain the priority corresponding to the call.

[0089] In one embodiment, the labeling unit 44 further includes:

[0090] The determination subunit is used to determine that the call is an invalid call when the similarity is less than or equal to a preset similarity threshold.

[0091] In one embodiment, the labeling unit 44 further includes:

[0092] The first prompting subunit is used to prompt the user to change the way of asking questions and interrupt the call after prompting a preset number of times.

[0093] In one embodiment, the first acquiring unit 41 includes:

[0094] A second judging subunit is used to judge whether the content sent online by the user is a question;

[0095] The second prompt sub-unit is used to prompt the user to send a consultation question when the content is not a question.

[0096] In one embodiment, the second prompt subunit includes:

[0097] A third judging subunit, configured to judge whether the content is a preset honorific when the content is not a question sentence;

[0098] The third prompting subunit is used to prompt the user to send a consultation question when the content is not a preset honorific.

[0099] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned seat service response processing device and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0100] Meanwhile, the division and connection method of each unit in the above-mentioned agent service response processing device is only used for illustration. In other embodiments, the agent service response processing device can be divided into different units as needed, and each unit in the agent service response processing device can also be connected in a different order and method to complete all or part of the functions of the above-mentioned agent service response processing device.

[0101] The above-mentioned seat service response processing device can be implemented in the form of a computer program. The computer program can be used in Figure 5 Runs on the computer device shown.

[0102] See also Figure 5 , Figure 5 1 is a schematic block diagram of a computer device provided in 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 in other devices.

[0103] See also Figure 5 The computer device 500 includes a processor 502, a memory and a network interface 505 connected via a system bus 501, wherein the memory may include a non-volatile storage medium 503 and an internal memory 504, and the memory may also be a volatile storage medium.

[0104] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. When the computer program 5032 is executed, the processor 502 can execute the above-mentioned agent service response processing method.

[0105] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0106] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute the above-mentioned agent service response processing method.

[0107] The network interface 505 is used to communicate with other devices over the network. Figure 5The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the computer device 500 to which the present application scheme is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or 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. In such an embodiment, the structure and function of the memory and the processor are the same as those of the present application scheme. Figure 5 The embodiments shown are consistent and will not be described again here.

[0108] The processor 502 is used to run a computer program 5032 stored in a memory to implement the following steps: respond to a call initiated by an online user and obtain a question contained in the call; automatically answer the question based on a preset knowledge graph to obtain interactive content, wherein the interactive content includes user content sent by the user side; match the user content with the content in the preset knowledge graph to obtain the similarity between the user content and the corresponding content in the preset knowledge graph; according to the similarity, prioritize the call and obtain the priority corresponding to the call; respond to the agent's instruction to obtain the call, and push the call to the agent in order of priority from high to low, so that the agent manually answers the call.

[0109] In one embodiment, when the processor 502 automatically answers the question based on the preset knowledge graph to obtain the interactive content, the processor 502 specifically implements the following steps:

[0110] Identify entities contained in the question to obtain user entities;

[0111] Matching the user entity with entities included in a preset knowledge graph;

[0112] In the case where there is a match between the entity contained in the preset knowledge graph and the user entity, obtaining the graph entity corresponding to the user entity in the preset knowledge graph, and linking the user entity to the graph entity;

[0113] Based on the link between the user entity and the graph entity, matching the question with a preset question associated with the graph entity;

[0114] When the question sentence matches the preset question sentence, a preset answer corresponding to the preset question sentence is responded to the user to obtain interactive content.

[0115] In one embodiment, when the processor 502 implements the priority marking of the call according to the similarity to obtain the priority corresponding to the call, the processor 502 specifically implements the following steps:

[0116] Determining whether the similarity is less than or equal to a preset similarity threshold;

[0117] When the similarity is greater than a preset similarity threshold, the call is marked with a priority to obtain the priority corresponding to the call.

[0118] In one embodiment, the processor 502 further implements the following steps:

[0119] When the similarity is less than or equal to a preset similarity threshold, the call is determined to be an invalid call.

[0120] In one embodiment, after determining that the call is an invalid call, the processor 502 further implements the following steps:

[0121] The user is prompted to change the way of asking questions, and after being prompted a preset number of times, the call is terminated.

[0122] In one embodiment, when obtaining the question sentence included in the call, the processor 502 specifically implements the following steps:

[0123] Determining whether the content sent online by the user is a question;

[0124] When the content is not a question, the user is prompted to send a question for consultation.

[0125] In one embodiment, when the processor 502 prompts the user to send a consultation question when the content is not a question, the processor 502 specifically implements the following steps:

[0126] In the case where the content is not a question sentence, determining whether the content is a preset honorific;

[0127] In the case that the content is not a preset honorific, the user is prompted to send a consultation question.

[0128] 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 (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0129] Those skilled in the art will appreciate that all or part of the processes in the method for implementing the above embodiment can be completed by a computer program, which can 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 process steps of the embodiment of the above method.

[0130] Therefore, 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, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the following steps:

[0131] A computer program product, when running on a computer, enables the computer to execute the steps of the seat service answering processing method described in the above embodiments.

[0132] The computer-readable storage medium may be an internal storage unit of the aforementioned device, such as a hard disk or 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 memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit of the device and an external storage device.

[0133] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0134] The storage medium is a physical, non-transient storage medium, for example, it can be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk, etc., which can store computer programs.

[0135] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0136] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0137] The steps in the method of the embodiment of the present application can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present application can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for an electronic device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0139] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A seat service answer processing method, characterized in that: include: Responding to a call initiated by a user online, and obtaining a question sentence included in the call; Based on a preset knowledge graph, the question is automatically answered to obtain interactive content, where the interactive content includes user content sent by the user side; Matching the user content with the content in the preset knowledge graph to obtain the similarity between the user content and the corresponding content in the preset knowledge graph; According to the similarity, the calls are marked with priorities to obtain the priorities corresponding to the calls; In response to the agent's instruction to obtain a call, the call is pushed to the agent in order of priority from high to low, and then the agent manually answers the call.

2. The seat service response processing method according to claim 1, characterized in that: Based on the preset knowledge graph, the question is automatically answered to obtain interactive content, including: Identify entities contained in the question to obtain user entities; Matching the user entity with entities included in a preset knowledge graph; In the case where there is a match between the entity contained in the preset knowledge graph and the user entity, obtaining the graph entity corresponding to the user entity in the preset knowledge graph, and linking the user entity to the graph entity; Based on the link between the user entity and the graph entity, matching the question with a preset question associated with the graph entity; When the question sentence matches the preset question sentence, a preset answer corresponding to the preset question sentence is responded to the user to obtain interactive content.

3. The seat service response processing method according to claim 1 or 2, characterized in that: According to the similarity, the call is marked with a priority to obtain the priority corresponding to the call, including: Determining whether the similarity is less than or equal to a preset similarity threshold; When the similarity is greater than a preset similarity threshold, the call is marked with a priority to obtain the priority corresponding to the call.

4. The seat service response processing method according to claim 3, characterized in that: The method further comprises: When the similarity is less than or equal to a preset similarity threshold, the call is determined to be an invalid call.

5. The seat service response processing method according to claim 4, characterized in that: After determining that the call is an invalid call, the method further includes: The user is prompted to change the way of asking questions, and after being prompted a preset number of times, the call is terminated.

6. The seat service response processing method according to claim 1, characterized in that: Get the questions contained in the call, including: Determining whether the content sent online by the user is a question; When the content is not a question, the user is prompted to send a question for consultation.

7. The seat service response processing method according to claim 6, characterized in that: In the case where the content is not a question, prompting the user to send a consultation question includes: In the case where the content is not a question sentence, determining whether the content is a preset honorific; In the case that the content is not a preset honorific, the user is prompted to send a consultation question.

8. A seat service answer processing device, characterized in that: include: A first acquisition unit, configured to respond to a call initiated by an online user and acquire a question sentence included in the call; An automatic answering unit, configured to automatically answer the question based on a preset knowledge graph to obtain interactive content, wherein the interactive content includes user content sent by a user side; A first matching unit, configured to match the user content with the content in the preset knowledge graph to obtain a similarity between the user content and the corresponding content in the preset knowledge graph; A labeling unit, configured to label the call with a priority according to the similarity, so as to obtain a priority corresponding to the call; The agent answering unit is used to respond to the agent's instruction to obtain a call, and push the call to the agent in order of priority from high to low, so that the agent manually answers the call.

9. A computer device, characterized in that: The computer device comprises a memory and a processor connected to the memory; the memory is used to store a computer program; the processor is used to run the computer program to execute the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 can be implemented.

Images