A method and system for automatically answering questions about business document data
By receiving questions in the chat group and assigning them to the business system server, and using preset templates to obtain third-party interface content, the problem of low efficiency in querying business documents is solved, and fast and accurate Q&A is achieved, improving the improvement of fast and accurate Q&A Work efficiency and system performance.
Patent Information
- Application Number
- CN202210462720.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-28
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2042-04-28
AI Technical Summary
In the prior art, the payment flow of business documents cannot be associated with the business order number, which makes it difficult for R&D personnel to provide accurate payment details in a timely manner, affecting work efficiency, and the query page is scattered, inconvenient inquiry, resulting in waste of resources and duplication problems.
Design a business document data automated Q&A method and system, by receiving question information in the chat group, sorting and assigning Q&A tasks to the corresponding business system server, using the preset Q&A template to call the third-party interface to obtain content, and generate the final Q&A content And send it back to the chat group to achieve automated Q&A.
It greatly improves the Q&A speed, reduces the workload of personnel, realizes server load balancing and hit rate of Q&A content, ensures the standardization and standardization of Q&A, and reduces the system workload.
Smart Images

Figure CN115062121B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of information technology, and specifically relates to a method and system for automatically answering questions about business document data. Background Art
[0002] Business documents have clearing and settlement business. Usually, the clearing process data is operated within the company, and all details are solidified, so that the calculation of the money for each document can be clearly seen through the details. However, the settlement process is to call the three-party payment interface to perform payment operations, but the three-party payment flow cannot be associated with the business order number, which makes it impossible for the document to perceive the payment details. When business personnel inquire about the payment situation, they cannot provide accurate payment details in time. R&D personnel need to obtain the data through the interface log or the three-party query page. In addition, the query page for querying the three-party data is not on the same page, but scattered in various systems, and the query page is not very convenient.
[0003] R&D personnel need to invest time and energy to answer these questions, which results in the distraction of R&D energy, interruption of current work, and impact on work efficiency. Different people may ask the same question multiple times, and the same person may ask the same question multiple times at different times. Summary of the invention
[0004] Technical problem: In view of the above problems existing in the prior art, the technical problem to be solved by the present invention is a method and system for automatically answering questions about business document data, which can meet the question-answering needs of different users, improve the efficiency of answering questions, and save time for answering questions.
[0005] Technical solution: In order to solve the above technical problems, the present invention provides a method for automatically answering questions about business document data, the method comprising:
[0006] Receive question information from questioners in the chat group;
[0007] Organize question information into Q&A tasks based on keywords in the question information, and automatically assign the Q&A tasks to the corresponding business system servers;
[0008] Pre-set different Q&A templates based on keywords in question information;
[0009] The task assigned to the business system server obtains the third-party interface required for Q&A according to the pre-set Q&A template, obtains the content required by the Q&A template by calling the third-party interface, and assembles the text and values of multiple third-party interfaces into the final Q&A content in the Q&A template;
[0010] Send the Q&A content to the chat group and remind the questioner.
[0011] Preferably, keywords are extracted from the question information and a list related to the keywords is sent to the chat group, and the questioner is reminded to make a selection, and the question type is classified according to the questioner's selection to narrow the scope of the answer.
[0012] Preferably, a question-answering robot is set up in the chat group and receives question information from the questioner according to the questioner's call.
[0013] Preferably, when the questioner summons the question-answering robot, the question-answering robot extracts the questioner's information while receiving the question information.
[0014] Preferably, the questioner's information includes the questioner's name in the chat group, and the questioner's identity and department are found in the database based on the questioner's information, and different question-answering templates are pre-set based on the questioner's identity and department.
[0015] The present invention also provides an automated question-answering system for business document data, comprising a question-answering robot, a question-answering task management module connected to the question-answering robot, a question-answering task executor module connected to the question-answering task management module, a question-answering list module and a question-answering template module connected to the question-answering task executor module;
[0016] The question answering robot is used to receive and send information in the chat group;
[0017] The question answering task management module receives question information sent by the question answering robot, organizes the question information into question answering tasks, and automatically assigns the question answering tasks to corresponding business system servers;
[0018] The Q&A task executor module is responsible for the execution of the Q&A task. It obtains the third-party interface required for Q&A through the Q&A template module for the task assigned to the business system server, obtains the content required by the Q&A template module by calling the third-party interface, and assembles the parameters configured by the Q&A template module into the Q&A content that is finally presented to the user.
[0019] The question-answering list module is used for the front-end visual display of answerable questions and the selection of the question-answering list;
[0020] The question-answering template module pre-sets different question-answering templates according to different question information, and the configuration of the question-answering template includes specific fields of the question-answering and corresponding values of the fields.
[0021] Preferably, the question answering task management module organizes the question information into different question answering tasks according to keywords in the question information, and distributes the question answering tasks to business system servers related to the keywords according to the keywords.
[0022] Preferably, the question-answering robot receives question information from a questioner when the chat group is summoned, and the question-answering robot extracts questioner information while receiving the question information.
[0023] Preferably, the question-answering template module also pre-sets different question-answering templates according to different questioner information, and the questioner information includes the name of the questioner in the chat group.
[0024] Preferably, it also includes a background management module, which is used for managing configurable items of the question-answering system.
[0025] Beneficial effects: Compared with the prior art, the present invention has the following advantages: 1. Automatic Q&A is achieved through the Q&A robot, and the inquiry process of the business personnel can be completed without intervention. The entire Q&A robot interaction process takes only a few seconds, which greatly improves the speed of Q&A and reduces the workload of personnel; 2. Q&A tasks are assigned according to question keywords, server load is balanced, and the hit rate of Q&A content is improved; 3. Q&A templates are set according to different questioning groups and group needs, so that they can be tailored to each person, and the Q&A content is standardized and normalized. There is no need to search and list all the information, which reduces the workload of the system. At the same time, due to the accurate configuration of the Q&A template, the hit rate of Q&A is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 It is a diagram of the question answering system of the present invention;
[0027] Figure 2 It is a schematic diagram of a payment information template in the question-answering template of the present invention;
[0028] Figure 3 It is a schematic diagram of a settlement information template in the question-answering template of the present invention;
[0029] Figure 4 It is a schematic diagram of an order information template in the question-answering template of the present invention;
[0030] Figure 5 It is a schematic diagram of a performance information template in the question-answering template of the present invention;
[0031] Figure 6 Schematic diagram of a chat group dialog box in an embodiment of the question answering method of the present invention;
[0032] Figure 7 2. It is a schematic diagram of the question-answering content of the robot embodiment of the present invention;
[0033] Figure 8 It is another schematic diagram of the configuration of the question-answering template of the present invention. DETAILED DESCRIPTION
[0034] The present invention is further illustrated below in conjunction with specific examples. The examples are implemented based on the technical solutions of the present invention. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention.
[0035] like Figure 1 As shown, a business document data automatic question-answering system includes a question-answering robot, a question-answering task management module, a question-answering task executor module, a question-answering list module, a question-answering template module and a background management module. Figure 1 The business system is the support system of the Q&A system (not part of the Q&A system). The corresponding instructions can be obtained through the Q&A logic to complete the ability of the system to call the business to obtain business data. For example, to obtain the business data of an order, it is necessary to obtain the business data by calling the three-party interface of the order. This process needs to be completed by implementing the set Q&A logic and instructions. The interactive communication technologies used at the bottom of this system are MQ (asynchronous interactive queue) and DUBBO (real-time interactive interface). The data storage technologies used at the bottom of this system are DB (MYSQL database) and REDIS (data cache).
[0036] In the embodiment of the present invention, the question-answering robot is used to receive and send information in a chat group; when someone in the chat group needs to ask a question, the questioner sends the question and summons the question-answering robot (i.e., @the robot in the chat group), the question-answering robot receives the summon and receives the question information, and the question-answering list and final question-answering content issued by the question-answering system are also sent by the question-answering robot and reminded by the question-answering robot (i.e., @the questioner in the chat group).
[0037] In the embodiment of the present invention, the question answering task management module receives question information sent by the question answering robot, organizes the question information into different question answering tasks according to the keywords in the question information, and assigns the question answering tasks to the business system servers related to the keywords according to the keywords. This function can automatically assign tasks to appropriate business system servers, and can achieve different needs for different people and load balancing. For example, if the keyword in the question information is "order" related information, the task is assigned to the order business system server to obtain order-related information; if the keyword in the question information is "inventory" related information, the task is assigned to the inventory business system server to obtain inventory-related information; if the keyword in the question information is "fulfillment" related information, the task is assigned to the fulfillment business system server to obtain fulfillment-related information; if the keyword in the question information is "settlement" related information, the task is assigned to the settlement business system server to obtain settlement-related information.
[0038] In the embodiment of the present invention, the Q&A task executor module is responsible for the execution of the Q&A task. The task assigned to the business system server obtains the third-party interface required for Q&A (i.e., the business data query interface provided by the business server) through the Q&A template module, obtains the content required by the Q&A template module by calling the third-party interface, and assembles the parameters configured by the Q&A template module into the Q&A content finally presented to the user. This execution module will also aggregate the content of multiple business system servers. For example, the entire life cycle of a business order may be scattered on different business system servers, and the Q&A content will aggregate the information of the entire order life cycle according to the template, such as order information, inventory information, fulfillment information, and settlement information.
[0039] In the embodiment of the present invention, the question and answer list module is used for the front-end visual display of questions that can be answered and the selection of the question and answer question list. The relevant question types of the questioner's question can be provided for the questioner to choose through a list selection method (sent to the chat group and @ the questioner by the question and answer robot), and the question types are preliminarily classified according to the questioner's choice. The classification of questions can effectively focus on the user's question and answer needs and accurately locate the question and answer keywords so that the system can better obtain the question and answer template and the question and answer content related to the accurate search keyword; at the same time, the order of the list will be dynamically displayed according to the number of times the question list is clicked, and the background will allocate more servers to answer such questions. For example, if the order business question has a high click-through rate, the question list will first display the order business question in the first place, and secondly, the background will allocate more server resources to process the order business question to ensure that the server resource allocation for question and answer processing is reasonable and efficient.
[0040] In the embodiment of the present invention, the question and answer template module pre-sets different question and answer templates according to different question information. The configuration of the question and answer template includes specific fields of the question and answer and the corresponding values of the fields. The template can standardize and normalize the content of the question and answer. The operation can set different question and answer templates according to different question information, set different question and answer templates according to keywords in the question information, and set different question and answer templates for different questioners and questioner needs. The question and answer robot extracts the questioner information while receiving the question information. The questioner information includes the name of the questioner in the chat group. The question and answer system is connected to the company's personnel information database, and the identity and department of the questioner are found in the database according to the name of the questioner in the chat group. Different question and answer templates are pre-set according to the identity and department of the questioner. If the personnel of the business department are concerned about business data, then the template can set relevant business fields. If the questioner is a department manager, then the template can set relevant business management fields. The Q&A template module includes multiple sub-template modules: payment information template module, settlement information template module, order information template module, fulfillment information template module, after-sales information template module and marketing information template module. Each sub-template contains multiple specific fields. When configuring the Q&A template, select one or more sub-templates, and when configuring each sub-template, select one or more specific fields. Figure 2 As shown, the Q&A template includes a payment information template. The specific fields included in the payment information template include: business order number, amount to be settled to the store, amount settled to the store, payment amount, account information, payment time, payment remarks, etc. Figure 3 As shown, the Q&A template also includes a settlement information template, which includes the following specific fields: settlement channel, settlement data status, settlement amount, settled amount, settlement status, settlement subject, settlement start time, and settlement time; Figure 4 As shown, the Q&A template also includes an order information template, which includes the following specific fields: order label, order payable amount, order actual payment amount, order status, order time, payment amount, payment status, payment time, and payment method; Figure 5 As shown, the Q&A template also includes a performance information template, which includes specific fields such as write-off status, performance amount, first write-off time, etc. Different Q&A templates are set according to the keywords in the question information, and different Q&A templates are set for different questioners and questioner needs. Different Q&A templates include different sub-templates, and different sub-templates have different specific fields, thus forming Q&A templates with different faces for different people.
[0041] The background management module in the embodiment of the present invention is mainly for the management of configurable items of the question-answering system, such as the addition, deletion, modification and query of question-answering duty group information and question-answering duty person information. This module is designed to enable operation and maintenance to better maintain the daily operation and maintenance of the question-answering system.
[0042] An embodiment of the present invention also discloses a method for automatically answering questions about business document data. The method includes:
[0043] S1. Set up a question-answering robot in the chat group and receive the question information of the questioner according to the summons of the questioner. When the questioner summons the question-answering robot, the question-answering robot extracts the information of the questioner while receiving the question information.
[0044] For example Figure 6 As shown, Xu * from the business department has the nickname "Xu * (Business)" in the chat group "Kangzhong Q&A Group 1". He enters "@Settlement Secretary, the order number is NCZ26491033839, and the store said it didn't receive the payment amount" in the chat group, wanting to obtain the order information about this order, and @ the question-answering robot (nickname in this group: Settlement Secretary). The question-answering robot receives the question information and simultaneously extracts the nickname "Xu * (Business)" of the questioner.
[0045] S2. Extract the keywords in the question information, send the list of relevant question types related to the keywords to the chat group, and remind the questioner to select. The list of relevant question types is to better focus on the question-answering needs of users, narrow the scope of question answering, and better generate question-answering tasks before generating the question-answering task.
[0046] In this embodiment: The question-answering system extracts the keywords "order" and "amount", and sends the list of question types related to these two keywords ● Tmall Auto Care / F6 roll-up / Installation service new link; ● Deposit refund; ● Localization / Large customers (such as Minsheng Bank) (as shown in the conversation in Figure 6 to the chat group through the question-answering robot Settlement Secretary and @ Xu *. Xu * selects "Tmall Auto Care / F6 roll-up / Installation service new link" in the list.
[0047] S3. Organize the question information into a question-answering task according to the keywords in the question information, and automatically assign the question-answering task to the corresponding business system server.
[0048] In this embodiment: The question-answering system obtains the question type, and according to the keywords "order" and "amount", obtains the business system servers "Order Business Processing Server" and "Settlement Business Processing Server".
[0049] S4. Find the identity and department of the questioner in the database according to the information of the questioner, and preset different question-answering templates according to the information of the questioner and the keywords in the question information.
[0050] In this embodiment: according to the name of the questioner "Xu* (business)" extracted by the Q&A robot in the chat group, the Q&A system finds the real name of the questioner in the company's personnel information database: Xu*, department: business department, and the core fields that users in the business department are concerned about include "amount to be settled at the store", "amount settled at the store", "account information", "amount of payment", "time of payment" and other fields. Then, according to the keywords "order" and "amount", since this question already has an order number, and the questioner is not concerned about the fields related to the fulfillment information for the time being, the Q&A templates are configured to include a payment information template, a settlement information template, and an order information template. The payment information template includes: "Business order number: ******", "amount to be settled at the store: ** yuan", "amount settled at the store: ** yuan", "amount of payment: ** yuan", "account information: ******", "time of payment" and so on. The fields in the settlement information template include: "Settlement channel: ******", "Settlement data status: ******", "Settlement amount: ** yuan", "Settled amount: ******", "Settlement status: ******", "Settlement subject: ******", "Start settlement time: ******", "Settlement time: ******", and the fields in the order information template include: "Order tag: ******", "Order payable amount: ** yuan", "Order actual payment amount: ** yuan", "Order status: ******", "Order time: ******", "Payment amount: ** yuan", "Payment status: ******", "Payment time: ******", and "Payment method: ******" (fields such as Figure 7 as shown).
[0051] S5. The task assigned to the business system server obtains the third-party interface required for answering questions according to the preset question answering template, obtains the content required by the question answering template by calling the third-party interface, and assembles the text and numerical values of multiple third-party interfaces into the final question answering content in the question answering template;
[0052] In this embodiment: The question type and order number NCZ26491033839 are used to call the interfaces of the "order business processing server" and the "settlement business processing server" to obtain relevant business information and settlement information. Through the payment information template, settlement information template, and order information template configured in the background, according to the relevant business information and settlement information returned by the configured field mapping interface. For example, in the configured order information template, the field is "actual paid amount of the order" and the code is: realPayAmount. The field name returned through this code mapping interface is "realPayAmount", and the corresponding value "99.00" of the field is placed in the configured order information template. For example, in the configured settlement information template, the field "settlement amount" has a code of: Settlementamount. The field name returned through this code mapping interface is "Settlement amount", and the corresponding value "98.62" of the field is placed in the configured settlement information template.
[0053] The Q&A system will splice the payment information template, settlement information template, and order information template with specific numerical values in the configured order. The "payment information" is placed at the front, followed by the "settlement information", and finally the "order information" (as Figure 7 shown).
[0054] S6. Send the Q&A content to the chat group and remind the questioner.
[0055] As Figure 7 shown, in this embodiment: The Q&A robot settlement secretary sends the Q&A content containing the payment information, settlement information, and order information of NCZ26491033839 to the chat group and @s Xu* (business) to complete the Q&A work.
[0056] Simplification of the Q&A content: In step S4, if the fields concerned by the department where the questioner is located are fewer, then a small amount of content is configured in the Q&A template. As Figure 8 shown, the department where the questioner is located only cares about specific fields such as "order payable amount", "actual paid amount of the order", "order status", and "order placement time" in the order information template, and specific fields such as "fulfillment status", "fulfillment amount", and "fulfillment time" in the fulfillment information template. Then, only the order information template and the fulfillment information template are configured in the Q&A template, and only the fields concerned by this department are selected for configuration in each template, thus greatly simplifying the process of finding the corresponding numerical values of specific fields, improving the Q&A speed, and simplifying the Q&A content.
[0057] The Q&A robot has handled a total of 740 questions in Kangzhong Q&A group. The robot answers an average of about 80 questions per week, with a hit rate of 98.67%. It has replaced the manual Q&A process in most scenarios, allowing developers to focus more on other things without interrupting what they are doing due to online Q&A issues, thereby improving R&D efficiency.
[0058] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. A method for automatically answering questions about business document data, characterized in that: The method comprises: A question-answering robot is set up in the chat group and receives the question information of the questioner according to the call of the questioner. When the questioner calls the question-answering robot, the question-answering robot extracts the questioner's information while receiving the question information. The questioner's information includes the name of the questioner in the chat group, and finds the identity and department of the questioner in the database according to the questioner's information; Extract keywords from question information and send a list related to the keywords to the chat group, and remind the questioner to choose. According to the questioner's choice, the question type is classified to narrow the scope of answering questions. Organize question information into Q&A tasks based on keywords in the question information, and automatically assign the Q&A tasks to the corresponding business system servers; Pre-set different Q&A templates based on keywords in the question information and the identity and department of the questioner; The task assigned to the business system server obtains the third-party interface required for Q&A according to the pre-set Q&A template, obtains the content required by the Q&A template by calling the third-party interface, and assembles the text and values of multiple third-party interfaces into the final Q&A content in the Q&A template; Send the Q&A content to the chat group and remind the questioner.
2. A business document data automated question answering system, characterized in that: It includes a question-answering robot, a question-answering task management module connected to the question-answering robot, a question-answering task executor module connected to the question-answering task management module, a question-answering list module and a question-answering template module connected to the question-answering task executor module; The question answering robot is used to receive and send information in the chat group; The question answering task management module receives question information sent by the question answering robot, organizes the question information into question answering tasks, and automatically assigns the question answering tasks to corresponding business system servers; The Q&A task executor module is responsible for the execution of the Q&A task. It obtains the third-party interface required for Q&A through the Q&A template module for the task assigned to the business system server, obtains the content required by the Q&A template module by calling the third-party interface, and assembles the parameters configured by the Q&A template module into the Q&A content that is finally presented to the user. The question-answering list module is used for the front-end visual display of answerable questions and the selection of the question-answering list; The question-answering template module pre-sets different question-answering templates according to different question information, and the configuration of the question-answering template includes specific fields of the question-answering and corresponding values of the fields.
3. The business document data automated question answering system according to claim 2, characterized in that: The question-answering task management module organizes question information into different question-answering tasks according to keywords in the question information, and distributes the question-answering tasks to business system servers related to the keywords according to the keywords.
4. The business document data automated question answering system according to claim 2, characterized in that: The question-answering robot receives question information from a questioner when the chat group is summoned, and extracts questioner information while receiving the question information.
5. The business document data automated question answering system according to claim 4, characterized in that: The question-answering template module also pre-sets different question-answering templates according to different questioner information, and the questioner information includes the name of the questioner in the chat group.
6. The business document data automated question answering system according to claim 2, characterized in that: It also includes a background management module, which is used for managing configurable items of the question-answering system.
