WO2020077874A1 - Method and apparatus for processing question-and-answer data, computer device, and storage medium - Google Patents

Abstract

A method for processing question-and-answer data, comprising: receiving user answer data and a user answer image sent by a terminal; determining, according to a preset determination method, whether the user answer data is correct; upon determining that the user answer data is correct, determining, according to a preset expression score determination method, a target expression score corresponding to the user answer image; determining target question data according to a preset question data determination method corresponding to the target expression score; and sending the target question data to the terminal for display.

Description

Question and answer data processing method, device, computer equipment and storage medium

This application requires priority to be submitted to the China Patent Office on October 16, 2018, with the application number 2018112354664, and the priority of the Chinese patent application titled "Question and Answer Data Processing Methods, Devices, Computer Equipment, and Storage Media." Incorporated in this application.

Technical field

This application relates to a question and answer data processing method, device, computer equipment, and storage medium.

Background technique

Questions and answers in traditional credit scenarios mostly rely on manual implementation. With the continuous development of artificial intelligence technology, intelligent question and answer based on systematic human-computer interaction gradually emerges. Terminals instead of salespersons ask users questions, reducing the industry for salespersons. Training costs.

However, the inventor realized that the currently implemented smart question answering method is relatively simple, usually a point-to-point verification method, that is, the server verifies the current smart question answering by determining the correctness of the user's answer data. In this way, the accuracy of the verification may be low, thereby reducing the accuracy of intelligent question answering, that is, for the question answering data in the process of intelligent question answering, there is a problem of low accuracy of question and answer data processing.

Summary of the invention

According to various embodiments disclosed in the present application, a question and answer data processing method, device, computer device, and storage medium are provided.

A question and answer data processing method includes:

Receive user answer data and user answer images sent by the terminal;

Determine whether the user's answer data is correct according to a preset determination method;

When it is determined that the user answer data is correct, the target expression score corresponding to the user answer image is determined according to a preset expression score determination method;

Determine the target question data according to the preset question data determination method corresponding to the target facial expression score; and

Send the target question data to the terminal for display.

A question and answer data processing device includes:

The receiving module is used to receive user answer data and user answer images sent by the terminal;

The determination module is used to determine whether the user answer data is correct according to a preset determination method;

A score determination module, configured to determine a target expression score corresponding to the user answering image according to a preset expression score determination method when the user answer data is determined to be correct;

A question data determination module, configured to determine target question data according to a preset question data determination method corresponding to the target expression score; and

The sending module is configured to send the target question data to the terminal for display.

A computer device includes a memory and one or more processors. The memory stores computer readable instructions. When the computer readable instructions are executed by the one or more processors, the one or more Each processor implements the steps of the question and answer data processing method provided in any embodiment of the present application.

One or more non-volatile computer-readable storage media storing computer-readable instructions, which when executed by one or more processors, cause the one or more processors to implement any of the present application The steps of the question and answer data processing device method provided in an embodiment.

The details of one or more embodiments of the application are set forth in the drawings and description below. Other features and advantages of this application will become apparent from the description, drawings, and claims.

BRIEF DESCRIPTION

In order to more clearly explain the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. Those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

FIG. 1 is an application scenario diagram of a question and answer data processing method according to one or more embodiments.

FIG. 2 is a schematic flowchart of a question and answer data processing method according to one or more embodiments.

FIG. 3 is a schematic flowchart of a method for processing question and answer data in another embodiment.

4 is a block diagram of a question and answer data processing device according to one or more embodiments.

5 is a block diagram of a question and answer data processing device in another embodiment.

Figure 6 is a block diagram of a computer device in accordance with one or more embodiments.

detailed description

In order to make the technical solutions and advantages of the present application more clear, the following describes the present application in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

The question and answer data processing method provided by this application can be applied to the application environment shown in FIG. 1. The terminal 102 communicates with the server 104 through the network through the network. The server 104 receives the user answer data and the user answer image sent by the terminal 102, and when the user answer data is determined to be correct according to the preset determination method, determines the target facial expression score corresponding to the user answer image, according to the preset corresponding to the target facial expression score The question data determination method determines the corresponding target question data, and sends the determined target question data to the terminal 102 for display. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server 104 may be implemented by an independent server or a server cluster composed of multiple servers.

In one of the embodiments, as shown in FIG. 2, a method for processing question and answer data is provided. Taking the method applied to the server in FIG. 1 as an example for illustration, it includes the following steps:

S202: Receive user answer data and user answer images sent by the terminal.

The user answer data is the user's answer data corresponding to the current question data. The user answer data may specifically be text data manually entered by the user on the terminal, or voice information collected by the terminal when the user answers the current question data. The user answering video is the image information correspondingly collected when the user responds to the current question data corresponding to the answer data. The user answer data may specifically be image information collected during the user answering process, for example, image information collected when the user manually inputs text data on the terminal for the current question data, or image information correspondingly collected when the user answers the current question data. The image information may specifically be information in the form of images or videos collected by the image collector. User answering images may include but are not limited to user answering images and user answering videos. The user answering image is, for example, an image corresponding to the user's face image collected when the user answers the question. The user answering video is, for example, the video collected during the user answering process, that is, the video collected during the user answering operation. The image collector may be a camera, and the camera may be a camera configured on the terminal, or may be an independent component connected point-to-point with the terminal. The user answer data corresponds to the user answer image.

The current question data is the current question data, that is, the question data corresponding to the current question and answer data processing process. The current question data may specifically refer to the question data that is closest to the current time, that is, the latest question data. Question data refers to the question data when the user asks a question. The question data may specifically be text data displayed through the display screen, or voice information displayed through voice broadcast. The current question data corresponds to the user answer data and the user answer image currently fed back by the terminal.

Specifically, the server receives user answer data and user answer images sent by the terminal through a wireless network or a wired network. The user answer data and the user answer image correspond to the current question data. When the terminal receives the current question data sent by the server, it displays the received current question data to the user through voice broadcast or display screen display, and detects in real time that the user responds to the user's answer data corresponding to the current question data. The terminal obtains the user answering image corresponding to the user answering data, and sends the acquired user answering image and the corresponding user answering data to the server.

In one of the embodiments, the server receives the question and answer instruction, obtains the corresponding current question data according to the received question and answer instruction, and sends the acquired current question data to the terminal. The server selects the current question data from the pre-stored preset question bank according to the question and answer instruction. The server may also obtain the corresponding question material from the pre-stored question material set according to the question and answer instruction, and generate corresponding current question data according to the acquired question material. The preset question bank is a question data set composed of multiple question data. The question material set is a material collection composed of multiple question materials. Question materials such as ID card number, nationality, current residence, etc.

In one of the embodiments, when the terminal presents the current question data to the user, the user's answer operation on the current question data is detected in real time, and the corresponding user answer data is obtained according to the detected answer operation. Answering operations such as user input operations on the terminal, or voice information generated when the user dictates answering data.

S204: Determine whether the user answer data is correct according to a preset judgment method.

The preset determination method is a preset method for determining whether the user answer data is correct. The preset judgment mode is the basis for judging whether the user answer data is correct. The preset determination method may be to compare the received user answer data with the pre-stored preset answers, and determine whether the user answer data is correct according to the comparison result, that is, determine the correctness of the user answer data. The preset answer is the preset correct answer or standard answer.

Specifically, the server queries the pre-stored preset answers corresponding to the current question data corresponding to the user answer data, matches the queried preset answer with the received user answer data, and determines the user answer according to the matching result Is the data correct? When the matching is successful, the server judges that the user answer data is correct; when the matching fails, it determines that the user answer data is wrong. The server can calculate the matching rate between the preset answer and the corresponding user answer data. When the calculated matching rate reaches the preset matching rate threshold, the server determines that the user answer data matches the corresponding preset answer successfully.

In one of the embodiments, when the user answer data is text data, the server matches the received text data with the corresponding preset query, and then determines the correctness of the text data. In one embodiment, when the user answers the data voice information, the server performs voice recognition on the voice information to obtain corresponding voice text content, and matches the recognized voice text content with the corresponding preset answer to determine the voice The correctness of the text content, and then determine the correctness of the corresponding voice information.

S206, when it is determined that the user answer data is correct, the target expression score corresponding to the user answer image is determined according to the preset expression score determination method.

The preset method for determining the facial expression score is a preset method for determining the target facial expression score corresponding to the user answering image. The preset expression determination method may specifically include extracting corresponding user micro-expressions from the user answering images, and determining a target expression score corresponding to the user answering images according to the extracted user micro-expressions by means of micro-expression recognition technology. The target facial expression score is the user facial expression score determined according to the user answer image corresponding to the user answer data. The target expression score may specifically be a user expression score determined according to the user's micro-expression contained in the user answering image. The target facial expression score can be used to characterize the likelihood that the user actually knows the correct answer corresponding to the current question data. The higher the target expression score, the greater the possibility that the user actually knows the correct answer. In other words, the target facial expression score may be determined according to the user's micro-expression, and the user actually knows the likelihood of the correct answer corresponding to the current question data. The target expression score can be used to determine whether the user answered the current question data because he actually knew the correct answer, or because he did not actually know the correct answer by guessing. The target expression score may refer to the score corresponding to the user's micro-expression, such as 80 points. The target expression score may also refer to the percentage corresponding to the user's micro-expression, such as 80%. The target expression score of 80% may refer to the possibility that the user actually knows the correct answer is 80%.

The user's micro-expression is a very short facial expression that cannot be controlled autonomously by humans. It is a very fast expression with a duration of only 1/25 second to 1/5 second. The user's micro-expressions cannot be hidden or disguised artificially, and can reflect the facial expressions of people's true emotions. According to the user's micro-expressions, it can be determined whether the user is straightforward or hesitant when answering the question.

Specifically, when it is determined that the user answer data is correct, the server extracts the corresponding user micro-expression from the user answer image corresponding to the user answer data, and uses the micro-expression recognition technology to determine the user's micro-expression according to the extracted user micro-expression The target expression score corresponding to the answer image. The server inputs the user answering image into the trained micro-expression recognition model for prediction, and obtains corresponding user micro-expressions. The server inputs the predicted micro-expression of the user into the trained expression score prediction model for prediction, and obtains the corresponding target expression score, that is, the target expression score corresponding to the user answering image.

In one of the embodiments, when it is determined that the user answer data is correct, the server inputs the corresponding user answer image into the trained prediction model for prediction, and obtains the corresponding target label score.

S208: Determine target question data according to a preset question data determination method corresponding to the target facial expression score.

The target question data is the question data the next time a question is initiated to the user. Target question data refers to another question data after the current question data. The target question data may specifically be text data that can be displayed through the display screen, or voice information that can be displayed through voice broadcast. The preset question data determination method is a preset method for determining the corresponding target question data according to the target expression score. The preset question data determination method corresponds to the target expression score. The method for determining the preset question data may specifically be to randomly select target question data from the preset question bank, or determine the corresponding target question data according to the current question data.

Specifically, the server determines the corresponding preset question data determination method according to the determined target expression score, and determines the corresponding target question data according to the determined preset question data determination method. The server pre-stores the correspondence between the target expression score and the preset question data determination method. The server queries the correspondence between the target facial expression score and the corresponding preset question data determination method according to the target facial expression score, and determines the preset question data determination method corresponding to the target facial expression score according to the queried correspondence . The server may select target question data of a preset type from the pre-stored preset question bank according to the way of determining the preset question data. The server may also determine the target question data related to the current question data according to the preset question data determination method, based on the current question data and the constructed knowledge graph.

In one embodiment, the server pre-stores the correspondence between the preset expression score interval and the preset question data determination method. The server matches the target expression score with the pre-stored preset expression score interval, and determines the preset question data determination method corresponding to the preset expression score interval that matches successfully as the pre-correspondence corresponding to the target expression score Set the question data determination method. The server matches the target expression score with each expression score in the preset expression score interval to determine whether the target expression score falls within the preset expression score interval, thereby determining the target expression score and Set the matching results between expression score intervals.

In one embodiment, the server may select target question data from a preset question bank. The server may also obtain corresponding question materials from the pre-stored question material sets, and generate corresponding target question data according to the obtained question materials. The target question data may or may not be related to the current question data. The question types corresponding to the target question data and the current question data may be the same or different. The question type corresponding to the target question data can be determined according to the question type corresponding to the current question data, the target expression score, and the corresponding preset question data determination method.

S210: Send target question data to the terminal for display.

Specifically, after determining the corresponding target question data according to the preset question data determination method corresponding to the target facial expression score, the server sends the determined target question data to the terminal. The terminal displays the received target question data to the corresponding user through display screen display or voice broadcast.

In one of the embodiments, when the terminal presents the received target question data to the user through voice announcement or display screen display, the user is detected in real-time to respond to the target question data corresponding to the user's answer data, and the detected user The answer data is sent to the server, so that the server continues to perform the relevant steps of the above question and answer data processing method according to the received user answer data.

In the above Q & A data processing method, the receiving terminal determines the correctness of the user ’s answer data according to the user ’s answer data and the user ’s answer image corresponding to the current question data, and executes the corresponding steps according to the correctness of the user ’s answer data to improve the processing of the Q & A data effectiveness. When it is determined that the user's answer data is correct, determine the target facial expression score corresponding to the user's answer image, determine the target question data for the next question according to the target facial expression score, and send the determined target question data to the terminal for display, In order to initiate the next question and answer process through the terminal. In this way, the target question answering data corresponding to the next question answering process is determined by combining the user answering data and the user answering image correspondence, thereby improving the processing accuracy of the question answering data in the current question answering process.

In one of the embodiments, the user answering image includes the user answering image; step S206 includes: when it is determined that the user answering data is correct, correspondingly acquiring the user answering image according to the user answering image; inputting the user answering image into a pre-trained micro-expression recognition model Make predictions to obtain corresponding user micro-expressions; enter user micro-expressions into a pre-trained expression score prediction model for prediction to obtain target expression scores.

The user answering image refers to an image including the user's face image correspondingly collected when the user answers the question. The user's face image refers to an image containing the user's face. The user's face image may specifically be the user's avatar photo or the user's headshot collected during the user's answering operation. The micro-expression recognition model is a model obtained by training the model based on the pre-acquired training sample set, and can be used to predict the micro-expression of an unknown user based on a known user answer image. The expression score prediction model is a model obtained by training the model based on the pre-acquired training sample set and can be used to predict unknown target expression scores based on known user micro-expressions.

Specifically, when it is determined that the user answer data is correct, the server acquires the corresponding user answer image from the user answer image corresponding to the user answer data. The server inputs the acquired user answer images as input features into a pre-trained micro expression recognition model, and predicts the input user answer images through the micro expression recognition model to obtain corresponding user micro expressions. The server inputs the user's micro-expressions predicted by the micro-expression recognition model into the pre-trained expression score prediction model, and predicts the user's micro-expressions through the expression score prediction model to obtain the corresponding target expression score.

In one of the embodiments, the server obtains a first training sample set, the first training sample set includes a target user answer image and a target user micro-expression corresponding to the target user answer image. There are multiple target user answer images, and each target user answer image corresponds to a corresponding target user micro-expression. The server uses each user answer image in the first training sample set as the input feature, and the corresponding target user's micro expressions as the desired output features, and trains the initial micro expression recognition model to obtain the trained micro expression recognition model. The server can obtain multiple target user answer images, and obtain corresponding target user answer images from each target user answer image.

In one of the embodiments, the server obtains a second training sample set, where the second training sample set includes the target user's micro expression and the target expression score corresponding to the target user's micro expression. There are multiple target user micro expressions, and each target user micro expression corresponds to a corresponding target expression score. The server takes each target user's micro-expressions in the second training sample set as input features, and the corresponding target expression scores as the desired output features, respectively, and trains the initialized expression score prediction model to obtain the trained expression scores Value prediction model. The server can obtain multiple target user answer images, and extract the corresponding target user micro expressions from each target user answer image through the trained micro expression recognition model.

In one of the embodiments, the user answering video includes multiple frames of user answering images. When it is determined that the user answer data is correct, the server obtains corresponding multi-frame user answer images from the user answer images. The server inputs each frame of user answer images to the trained micro-expression recognition model for prediction to obtain corresponding user micro-expressions, and enters the predicted user micro-expressions into the trained expression score prediction model for prediction to obtain corresponding Emoji score. The server correspondingly determines the target expression score corresponding to the corresponding user answer image according to the expression score corresponding to each frame of the user answer image. For example, the server may perform weighted evaluation on the expression score corresponding to each frame of the user answering image to obtain the corresponding target expression score.

In the above embodiments, with the aid of the trained micro-expression recognition model and expression score prediction model, the corresponding target expression score is obtained according to the user answer image included in the user answer image, which improves the prediction efficiency and accuracy of the target expression score To improve the efficiency and accuracy of question and answer data processing.

In one of the embodiments, the user answering image includes the user answering video; step S206 includes: when it is determined that the user answering data is correct, correspondingly obtaining the user answering video according to the user answering image; extracting the preset from the user answering video according to the preset extraction method Number of video frames; separately determine the expression score corresponding to each video frame; according to each expression score, determine the target expression score corresponding to the user answering image.

The user answering video refers to the correspondingly collected video during the user's answering process, that is, the video collected during the user's answering operation. Video frames are the basic unit of video for users to answer questions. One video frame corresponds to one static picture in the video, and multiple video frames make up the video. The preset number is a preset value, which can be customized according to the actual situation, such as 3. The preset extraction method is a preset method for extracting a preset number of video frames from the user answer video. The preset extraction method is used to instruct the server how to obtain a preset number of video frames from the user answer video. The preset extraction method may be to obtain a preset number of consecutive video frames from the user answering video, or to obtain a preset number of video frames from the user answering video according to a preset step.

Specifically, when it is determined that the user answer data is correct, the server acquires the corresponding user answer video from the user answer image corresponding to the user answer data. The server extracts a preset number of video frames from the obtained user answer video according to the preset extraction method, and uses the trained micro-expression recognition model and expression score prediction model to perform each video frame separately in the above manner Predict, get the expression score corresponding to each video frame. The server determines the target expression score corresponding to the corresponding user answering image according to the corresponding expression scores of the preset number of video frames. The server may perform weighted evaluation on the expression score corresponding to each video frame to obtain the corresponding target expression score. The server may also use the expression score corresponding to each video frame as an input feature, input a pre-trained score prediction model for prediction, and obtain a corresponding target expression score.

In one of the embodiments, the server obtains a preset number of consecutive video frames from the user answer video. In one of the embodiments, the server obtains a preset number of video frames from the user answer video according to a preset step. The interval between any two adjacent video frames in the preset number of acquired video frames is the preset step size. The preset step size refers to the interval between two adjacent video frames. The preset step size may specifically be a specified number of video frames, such as 5 frames. The server acquires a video frame every specified number of video frames in the user answer video, and stops acquiring until a preset number of video frames are acquired.

In the above embodiment, the server extracts a preset number of video frames from the user answering video, and integrates the corresponding expression scores of the preset number of video frames to determine the corresponding target expression score, which improves the accuracy of the target expression score, thereby improving The accuracy of question and answer data processing.

In one of the embodiments, step S208 includes: comparing the target expression score with the preset expression score; when the target expression score reaches the preset expression score, selecting the target of the preset topic type from the preset question bank Ask the data.

The preset expression score is a preset expression score threshold, such as 86 points, or 86%. The preset expression score is the basis for determining the corresponding preset question data determination method according to the target expression score. The preset question type is a preset question type. The question type refers to the type corresponding to the question data, such as identity type and address type. Identity type question data such as "What is the last six digits of your ID number?" Address type question data such as "What is your current residential address"

Specifically, the server compares the determined target facial expression score with the preset facial expression score. When the target expression score reaches the preset expression score, it indicates that the user's expression during the answering process is normal, which means that the user can determine that the user actually knows the correct answer to the current question data according to the user's answer image, which further verifies the corresponding user answer data Correctness, the server selects the target question data of the preset question type from the pre-stored preset question bank.

In one of the embodiments, the server is locally pre-configured with one or more question types corresponding to the question and answer session. For multiple pre-configured question types, the server is locally pre-configured with the question and answer sequence corresponding to the corresponding question and answer process for the multiple question types, such as the address type question data followed by the identity type question data. Assuming that the question type corresponding to the current question data is the address type, when the target expression score corresponding to the current question data reaches the preset expression score, the server determines that the question type corresponding to the target question data is the identity type according to the pre-configured question and answer sequence . The server filters the question data of the identity type from the preset question bank, and selects the target question data from the filtered question data.

In one of the embodiments, the preset question type corresponding to the target question data is different from the question type corresponding to the current question data.

In the above embodiment, when the user answer data corresponding to the current question data is determined to be correct according to the user answer data and the user answer image, the question data of other preset question types is selected to continue to ask the user to reduce the number of question data , So as to reduce the number of question and answer data processing in the entire question and answer process, thereby improving the efficiency of question and answer data processing.

In one of the embodiments, the user answer data and the user answer image correspond to the current question data; step S208 includes: comparing the target expression score with the preset expression score; when the target expression score is lower than the preset expression score At this time, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data.

The preset selection method is a preset method for selecting target question data according to the current question data. The preset selection method may specifically be to select target question data related to the current question data from the preset question bank, or to select corresponding target question data based on the constructed knowledge graph. The question type corresponding to the target question data selected according to the preset selection method may be the same as or different from the question type corresponding to the current question data.

Specifically, the server compares the target facial expression score corresponding to the user answering image with the preset facial expression score. When the target expression score is lower than the preset expression score, the server selects the target question data related to the current question data according to the current question data corresponding to the user answer data and the user answer image according to the preset selection method. The server may select target question data related to the current question data from the preset question bank according to the current question data. The server may also select relevant target question data from the constructed knowledge graph according to the current question data. The current question data and the correspondingly determined target question data may correspond to the same or corresponding keywords, such as an address that is the same or similar.

In one embodiment, when the target expression score is lower than the preset expression score, the server selects target question data related to the user answer data according to the preset selection method according to the user answer data corresponding to the current question data. For example, suppose the current question data is "What is your current residential address", and the corresponding user answer data is "I currently live in No. 28 Keyuan Road, Nanshan District, Shenzhen". When the target expression score is lower than the preset expression score, the server may select the corresponding target question data according to the address "28 Keyuan Road, Nanshan District, Shenzhen" in the user answer data, for example, "From Shenzhen Nanshan Branch There is a China Merchants Bank within 50 meters of No. 28 Yuan Road, right? "

In one of the embodiments, the preset question bank is a question bank pre-configured into the database. The constructed knowledge graph is a graph-based data structure constructed in advance according to the knowledge graph construction method. The construction of the knowledge graph can be realized based on a known way of constructing the knowledge graph, which will not be repeated here.

In the above embodiment, when it is determined that the user's answer data is correct, and the correctness of the user's answer data is determined based on the corresponding user's answer image, the target question data related to the current question data is selected to further determine the current question based on the target question data The accuracy of the data, thereby verifying the reliability of the identity of the user currently answering the question, and ensuring the correctness of the question and answer data processing.

In one of the embodiments, the user answer data and the user answer image correspond to the current question data; the current question data corresponds to a preset comprehensive score; the question and answer data processing method further includes: when the target expression score is lower than the preset expression score When the value is set, the preset comprehensive score is dynamically adjusted according to the preset score adjustment method; according to the preset comprehensive score and the adjusted preset comprehensive score, the comprehensive score corresponding to the target question data is determined.

The preset comprehensive score is a preset comprehensive score corresponding to the current question data. The preset comprehensive score includes a preset expression score and a preset answer score. The preset comprehensive score can be understood as the full score corresponding to the current question data. The preset answer score can be understood as the full score corresponding to the user ’s answer data. For example, when the user ’s answer data for the current question data is consistent with the correct answer, the corresponding target answer data will be determined as the full score. The preset answer score is determined as the target answer score. Similarly, the preset expression score value can be understood as the full score value corresponding to the user answering image.

Specifically, when the target expression score is lower than the preset expression score, the server dynamically adjusts the preset comprehensive score corresponding to the current question data corresponding to the user answer data and the user answer image according to the preset score adjustment method. The server determines the comprehensive score corresponding to the target question data according to the adjusted preset comprehensive score and the preset comprehensive score corresponding to the current question data. When it is determined that the correctness of the corresponding user's answer data is in doubt based on the user's answer image, the server dynamically transfers a part of the preset comprehensive score corresponding to the current question data to the target question data.

In one of the embodiments, the preset comprehensive score includes a preset expression score and a preset answer score. When the target expression score is lower than the preset expression score, the server determines the preset expression score corresponding to the current question data as the comprehensive score corresponding to the target question data. At the same time, the server determines the preset answer score corresponding to the current question data as the adjusted preset comprehensive score. The server may determine, according to the preset comprehensive score corresponding to the current question data, the percentage of the score corresponding to the preset expression score and the preset answer score, and correspondingly determine the expression score in the comprehensive score corresponding to the target question data Value and answer score.

In one of the embodiments, when the target expression score is lower than the preset expression score, the server determines the corresponding target answer score according to the user answer data, and determines the corresponding target synthesis according to the target answer score and the target expression score Score, and then the determined target comprehensive score is determined as the comprehensive score corresponding to the target question data.

In one of the embodiments, when the target expression score is lower than the preset expression score, the server determines the current question data and the target question data according to the preset score ratio and the preset comprehensive score corresponding to the current question data, respectively The corresponding comprehensive score after dynamic adjustment. For example, it is assumed that the preset score ratio is 3: 2, that is, after performing the dynamic adjustment step, the ratio of the comprehensive scores corresponding to the current question data and the target question data is 3: 2. Assuming that the preset comprehensive score is 50 points, the preset adjusted comprehensive score corresponding to the current question data corresponding to the adjusted adjusted score according to the preset score ratio is 30 points, and the comprehensive score corresponding to the target question data is 20 points.

In one of the embodiments, the server is locally pre-configured with the number of question data involved in the question and answer session and the preset comprehensive score corresponding to each question data.

In one of the embodiments, the server is locally pre-configured with multiple question types corresponding to the question and answer session, and a preset comprehensive score corresponding to each question type. The preset comprehensive score corresponding to each topic type may be the same or different. For the types of questions in the compulsory answer category, a higher preset comprehensive score can be configured, and the types of questions in the compulsory answer category such as the identity category. In the question and answer session, when the question data of a certain question type is selected for the first time, the server determines the preset comprehensive score corresponding to the question type as the preset comprehensive score corresponding to the question data first selected for the question type.

In the above embodiment, when it is determined that the correctness of the corresponding user ’s answer data is doubtful according to the user ’s answer image, the pre-synthetic score of the current question data is dynamically adjusted, and the comprehensive score corresponding to the target question data is correspondingly determined to facilitate questioning according to the target The target comprehensive score corresponding to each of the data and the current question data determines the final target comprehensive score corresponding to the current question data, improves the accuracy of the target comprehensive score, and thus improves the accuracy of question and answer data processing.

In one of the embodiments, before step S208, the above question and answer data processing method further includes: determining a target answer question score corresponding to the user answer data according to a preset answer question score determination method; according to the target answer question score and the target expression score Determine the corresponding target comprehensive score; update the existing total answer score according to the target comprehensive score; when the updated total answer score meets the preset stop condition, stop the current question and answer process and send the corresponding prompt message to the terminal .

The preset answer score determination method is a predetermined method to determine the target answer score according to the user answer data. The method for determining the preset answer score may be to determine the corresponding target answer score according to the matching rate between the user answer data and the corresponding preset answer, and the preset answer score corresponding to the current question data. For example, assuming that the preset comprehensive score is 50 points and the calculated matching rate is 80%, the corresponding target answer score is 40 points. The preset answer score determination method may also be to input user answer data into a pre-trained answer score prediction model for prediction to obtain a corresponding target answer score.

The existing total score for answering questions refers to the total score determined according to the target comprehensive score corresponding to one or more question data before the current question data. The target comprehensive score corresponding to the one or more question data and the target comprehensive score corresponding to the current question data correspond to the same user ID. In other words, the target comprehensive scores corresponding to the current question data and the one or more question data between the current question data are based on the user answer data and the user answer images corresponding to the corresponding user's corresponding question data. Definite answer score. The total answer score may be a total score determined according to the target comprehensive score corresponding to each question data. The total score of the answer can be the direct summation of multiple target comprehensive scores, or the total score obtained by weighted summation. The preset stop condition is a preset basis for determining whether to stop the current question and answer process. The preset stop condition may specifically be that the updated total answer score reaches the preset total score threshold.

Specifically, the server determines the corresponding target answer score according to the user answer data according to the preset answer score determination method, and sums the target answer score and the corresponding target facial expression score to obtain the corresponding target comprehensive score. The server queries the total score of the existing answers according to the user answer data and the user ID corresponding to the user's answer image, and correspondingly updates the query total score according to the target comprehensive score corresponding to the current question data. The server compares the updated total answer score with the preset total score threshold. When the updated total answer score reaches the preset total score threshold, the server stops the current question and answer process and does not continue to perform the relevant steps of determining the target question data according to the preset question data determination method corresponding to the target facial expression score. The server generates corresponding prompt information according to the updated total score of the answer and sends it to the terminal.

For example, assuming that the current question data is the third question data answered by the corresponding user, the target comprehensive score corresponding to the third question data determined in the above manner is 20 points, the first question data and the second question The target comprehensive scores corresponding to the data are 30 points and 20 points respectively, then the total score of the existing answer questions is 50 points, and the corresponding total score of the updated answer questions is 70 points. Assuming that the preset total score threshold is 60 points, the updated total answer score is greater than the preset total score threshold and meets the preset stop condition.

In one of the embodiments, the server directly sums or weights the target answer score and the target facial expression score to obtain the corresponding target comprehensive score. The weight of the weighted summation can be customized.

In one of the embodiments, when the question and answer data processing method in each of the above embodiments is applied to the question and answer link in the face-to-face loan review process, the server stops the corresponding question and answer process and pushes prompt information indicating that the face-to-face review is passed to the terminal. Interview refers to the verification of the user's identity in the process of lending business.

In one of the embodiments, the server counts the total number of question data, and compares the total number of question data with a preset threshold. When the total number of question data counted reaches the preset threshold, the server stops the current question and answer process, and updates the existing total answer score according to the target comprehensive score corresponding to the current question data, and then based on the updated answer question The total score value and the preset total score threshold push corresponding prompt information to the terminal. When the updated total score of the answers reaches the preset total score threshold, the server sends a prompt message indicating that the interview is passed to the terminal. When the updated total score of the answer is lower than the preset total score threshold, the server sends a prompt message to the terminal indicating that the interview has failed.

In one of the embodiments, when it is determined that the user answer data is correct, the server may determine the corresponding preset answer score as the target comprehensive score. When it is determined that the user answer data is wrong, the server may determine the corresponding target answer score to be zero.

In one of the embodiments, when it is determined that the user answer data corresponding to the current question data is wrong, the server may stop the current question and answer process and send corresponding prompt information to the terminal. The server may also select target question data of the preset question type from the preset question bank, and continue to execute the current question and answer process according to the selected target question data in the above manner. The server counts the total number of question data corresponding to the wrong answer data of the corresponding user. When the counted total number reaches the preset total number, the server stops the current question and answer process and sends corresponding prompt information to the terminal.

In one of the embodiments, the server receives the user answering voice information sent by the terminal, performs voiceprint recognition on the user answering voice information to obtain corresponding voiceprint features, and further verifies the identity of the corresponding user according to the voiceprint features. Voiceprint recognition can be implemented based on various existing voiceprint recognition technologies, and will not be repeated here. When it is determined according to the user answer data and the user answer image that the corresponding user's answer data for the current question data is correct, the server obtains the corresponding user answer voice information from the terminal.

In one of the embodiments, the server obtains the user's tone and intonation from the user's answering voice information, and further verifies the correctness of the user's answer data according to the obtained tone and intonation.

As shown in FIG. 3, in one of the embodiments, a method for processing question and answer data is provided. The method specifically includes the following steps:

S302: Receive user answer data and user answer images sent by the terminal.

S304: Determine whether the user answer data is correct according to a preset judgment method.

S306, when it is determined that the user answer data is correct, the user answer image is correspondingly obtained according to the user answer image.

S308, input the user answering image into a pre-trained micro-expression recognition model for prediction, and obtain corresponding user micro-expressions.

S310. Input a user's micro-expression into a pre-trained expression score prediction model for prediction to obtain a target expression score.

S312, when it is determined that the user answer data is correct, the user answer video is correspondingly obtained according to the user answer image.

S314: Extract a preset number of video frames from the user answer video according to the preset extraction method.

S316, separately determining the expression score corresponding to each video frame.

S318, according to each expression score, a target expression score corresponding to the user answering image is determined.

S320: Determine the target answer score corresponding to the user answer data according to the preset answer score determination method.

S322: Determine the corresponding comprehensive target score according to the target answer score and the target facial expression score.

S324: Correspondingly update the existing total score of the answer according to the target comprehensive score.

S326, when the updated total score of answering questions meets the preset stop condition, the current question and answer process is stopped, and corresponding prompt information is sent to the terminal.

S328: Compare the target expression score with the preset expression score.

S330: When the target expression score reaches the preset expression score, select target question data of the preset question type from the preset question bank.

S332, when the target expression score is lower than the preset expression score, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data; the user answer data and the user answer image correspond to the current question data.

S334: Send the target question data to the terminal for display.

In the above embodiment, the correctness of the corresponding user's answer data is assisted by the help of the user's answer image. When the user's answer data is verified to be correct, the corresponding target question data is determined according to the target expression score corresponding to the user's answer image, which improves the question and answer data Processing accuracy. Further, it is determined whether to stop the current question and answer process according to the total score of the questions answered, which improves the efficiency of question and answer data processing.

It should be understood that although the steps in the flowcharts of FIGS. 2-3 are displayed in order according to the arrows, the steps are not necessarily executed in the order indicated by the arrows. Unless clearly stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in FIGS. 2-3 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed and completed at the same time, but may be executed at different times. These sub-steps or stages The execution order of is not necessarily sequential, but may be executed in turn or alternately with at least a part of other steps or sub-steps or stages of other steps.

In one of the embodiments, as shown in FIG. 4, a question and answer data processing apparatus 400 is provided, including: a receiving module 402, a determining module 404, a score determining module 406, a question data determining module 408, and a sending module 410, wherein :

The receiving module 402 is used to receive user answer data and user answer images sent by the terminal.

The determination module 404 is used to determine whether the user answer data is correct according to a preset determination method.

The score determination module 406 is used to determine the target facial expression score corresponding to the user's answer image when it is determined that the user's answer data is correct.

The question data determination module 408 is configured to determine target question data according to a preset question data determination method corresponding to the target expression score.

The sending module 410 is used for sending target question data to the terminal for display.

In one of the embodiments, the user answer image includes the user answer image; the score determination module 406 is also used to obtain the user answer image corresponding to the user answer image when it is determined that the user answer data is correct; input the user answer image into the pre-training The prediction model of the micro-expression recognition model is used to obtain the corresponding micro-expression of the user; the micro-expression of the user is input into the pre-trained expression score prediction model for prediction to obtain the target expression score.

In one of the embodiments, the user answer video includes the user answer video; the score determination module 406 is also used to obtain the user answer video according to the user answer image when the user answer data is determined to be correct; Extract a preset number of video frames from the video; determine the expression score corresponding to each video frame; and determine the target expression score corresponding to the user's answer image according to each expression score.

In one of the embodiments, the question data determination module 408 is also used to compare the target expression score with the preset expression score; when the target expression score reaches the preset expression score, the pre-selected from the preset question bank Set the target question data for the question type.

In one embodiment, the user answer data and the user answer image correspond to the current question data; the question data determination module 408 is also used to compare the target expression score with the preset expression score; when the target expression score is lower than When the expression score is preset, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data.

As shown in FIG. 5, in one of the embodiments, the user answer data and the user answer image correspond to the current question data; the current question data corresponds to a preset comprehensive score; the question and answer data processing device 400 further includes: score adjustment Module 412;

The score adjustment module 412 is used to dynamically adjust the preset comprehensive score according to the preset score adjustment method when the target expression score is lower than the preset expression score; according to the preset comprehensive score and the adjusted preset synthesis Score, determine the comprehensive score corresponding to the target question data.

In one of the embodiments, the question and answer data processing device 400 further includes: a total score update module 414;

The total score update module 414 is used to determine the target answer score corresponding to the user answer data according to the preset answer score determination method; determine the corresponding target comprehensive score according to the target answer score and the target facial expression score; according to the target The integrated score corresponds to the update of the existing total score of answering questions; when the updated total score of answering questions meets the preset stop condition, the current question and answer process is stopped and corresponding prompt information is sent to the terminal.

For the specific limitation of the question and answer data processing device, please refer to the above limitation on the question and answer data processing method, which will not be repeated here. Each module in the above question and answer data processing device may be implemented in whole or in part by software, hardware, or a combination thereof. The above modules may be embedded in the hardware or independent of the processor in the computer device, or may be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one of the embodiments, a computer device is provided, the computer device may be a server, and the internal structure diagram thereof may be as shown in FIG. 6. The computer device includes a processor, memory, network interface, and database connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer-readable instructions, and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile storage medium. The database of the computer device is used to store preset answers corresponding to current question data corresponding to user answer data and user answer images. The network interface of the computer device is used to communicate with external terminals through a network connection. The computer readable instructions are executed by the processor to implement a question and answer data processing method.

Those skilled in the art can understand that the structure shown in FIG. 6 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may Include more or less components than shown in the figure, or combine certain components, or have a different arrangement of components.

A computer device includes a memory and one or more processors. The memory stores computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors implement any one of the applications The steps of the question and answer data processing method provided in the embodiments.

One or more non-volatile computer-readable storage media storing computer-readable instructions, which when executed by one or more processors, cause the one or more processors to implement any one of the embodiments of the present application The steps of the question and answer data processing method provided.

A person of ordinary skill in the art may understand that all or part of the process in the method of the foregoing embodiments may be completed by instructing relevant hardware through computer-readable instructions, and the computer-readable instructions may be stored in a non-volatile computer In the readable storage medium, when the computer-readable instructions are executed, they may include the processes of the foregoing method embodiments. Any references to memory, storage, databases, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be arbitrarily combined. To simplify the description, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, all It is considered as the scope described in this specification.

The above-mentioned embodiments only express several implementations of the present application, and their descriptions are more specific and detailed, but they should not be understood as limiting the scope of the invention patent. It should be noted that, for those of ordinary skill in the art, without departing from the concept of the present application, a number of modifications and improvements can also be made, which all fall within the protection scope of the present application. Therefore, the protection scope of the patent of this application shall be subject to the appended claims.

Claims (20)

1. A question and answer data processing method, including:

   Receive user answer data and user answer images sent by the terminal;

   Determine whether the user's answer data is correct according to a preset determination method;

   When it is determined that the user answer data is correct, the target expression score corresponding to the user answer image is determined according to a preset expression score determination method;

   Determine the target question data according to the preset question data determination method corresponding to the target facial expression score; and

   Send the target question data to the terminal for display.
2. The method according to claim 1, wherein the user answering image includes a user answering image; and when it is determined that the user answering data is correct, the answering image with the user is determined according to a preset expression score determination method The corresponding target expression scores include:

   When it is determined that the user answer data is correct, correspondingly obtain the user answer image according to the user answer image;

   Input the user answering image into a pre-trained micro-expression recognition model for prediction, and obtain corresponding user micro-expressions; and

   The user's micro-expression is input into a pre-trained expression score prediction model for prediction to obtain a target expression score.
3. The method according to claim 1, wherein the user answering image includes a user answering video; and when it is determined that the user answering data is correct, the answering image with the user is determined according to a preset expression score determination method The corresponding target expression scores include:

   When it is determined that the user answer data is correct, correspondingly obtain the user answer video according to the user answer image;

   Extract a preset number of video frames from the user answer video according to a preset extraction method;

   Separately determine the expression score corresponding to each of the video frames; and

   A target expression score corresponding to the user answering image is determined according to each of the expression scores.
4. The method according to claim 1, wherein the determining the target question data according to the preset question data determination method corresponding to the target expression score includes:

   Comparing the target expression score with a preset expression score; and

   When the target expression score reaches the preset expression score, the target question data of the preset question type is selected from the preset question bank.
5. The method according to claim 1, wherein the user answer data and the user answer image correspond to current question data; and the target question is determined according to the preset question data determination method corresponding to the target expression score Data, including:

   Comparing the target expression score with a preset expression score; and

   When the target expression score is lower than the preset expression score, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data.
6. The method according to any one of claims 1 to 5, wherein the user answer data and the user answer image correspond to current question data; the current question data corresponds to a preset comprehensive score; the The method also includes:

   When the target expression score is lower than the preset expression score, dynamically adjust the preset comprehensive score according to the preset score adjustment method; and

   According to the preset comprehensive score and the adjusted preset comprehensive score, the comprehensive score corresponding to the target question data is determined.
7. The method according to any one of claims 1 to 5, wherein before determining the target question data according to the preset question data determination method corresponding to the target expression score, the method further comprises:

   Determine the target answer score corresponding to the user answer data according to the preset answer score determination method;

   Determine a corresponding target comprehensive score according to the target answer score and the target facial expression score;

   Correspondingly update the existing total score of the answers according to the target comprehensive score; and

   When the updated total answer score meets the preset stop condition, the current question and answer process is stopped, and corresponding prompt information is sent to the terminal.
8. A question and answer data processing device, including:

   The receiving module is used to receive user answer data and user answer images sent by the terminal;

   The determination module is used to determine whether the user answer data is correct according to a preset determination method;

   A score determination module, configured to determine a target expression score corresponding to the user answering image according to a preset expression score determination method when the user answer data is determined to be correct;

   A question data determination module, configured to determine target question data according to a preset question data determination method corresponding to the target expression score; and

   The sending module is configured to send the target question data to the terminal for display.
9. A computer device includes a memory and one or more processors. The memory stores computer readable instructions. When the computer readable instructions are executed by the one or more processors, the one or more Each processor performs the following steps:

   Receive user answer data and user answer images sent by the terminal;

   Determine whether the user's answer data is correct according to a preset determination method;

   When it is determined that the user answer data is correct, the target expression score corresponding to the user answer image is determined according to a preset expression score determination method;

   Determine the target question data according to the preset question data determination method corresponding to the target facial expression score; and

   Send the target question data to the terminal for display.
10. The computer device according to claim 9, wherein the user answer image includes a user answer image; and when it is determined that the user answer data is correct, the answer to the user is determined according to a preset expression score determination method The target expression score corresponding to the image, including:

    When it is determined that the user answer data is correct, correspondingly obtain the user answer image according to the user answer image;

    Input the user answering image into a pre-trained micro-expression recognition model for prediction, and obtain corresponding user micro-expressions; and

    The user's micro-expression is input into a pre-trained expression score prediction model for prediction to obtain a target expression score.
11. The computer device according to claim 9, wherein the user answering image includes a user answering video; and when it is determined that the user answering data is correct, the answering to the user is determined according to a preset expression score determination method The target expression score corresponding to the image, including:

    When it is determined that the user answer data is correct, correspondingly obtain the user answer video according to the user answer image;

    Extract a preset number of video frames from the user answer video according to a preset extraction method;

    Separately determine the expression score corresponding to each of the video frames; and

    A target expression score corresponding to the user answering image is determined according to each of the expression scores.
12. The computer device according to claim 9, wherein the determining the target question data according to the preset question data determination method corresponding to the target expression score includes:

    Comparing the target expression score with a preset expression score; and

    When the target expression score reaches the preset expression score, the target question data of the preset question type is selected from the preset question bank.
13. The computer device according to claim 9, wherein the user answer data and the user answer image correspond to current question data; and the target is determined according to the preset question data determination method corresponding to the target facial expression score Question data, including:

    Comparing the target expression score with a preset expression score; and

    When the target expression score is lower than the preset expression score, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data.
14. The computer device according to any one of claims 9 to 13, wherein the user answer data and the user answer image correspond to current question data; the current question data corresponds to a preset comprehensive score; The processor also executes the following steps when executing the computer-readable instructions:

    When the target expression score is lower than the preset expression score, dynamically adjust the preset comprehensive score according to the preset score adjustment method; and

    According to the preset comprehensive score and the adjusted preset comprehensive score, the comprehensive score corresponding to the target question data is determined.
15. The computer device according to any one of claims 9 to 13, wherein when the computer-readable instructions are executed by the processor, the processor is caused to perform the correspondence according to the target expression score Before determining the default question data determination method, the following steps are also performed:

    Determine the target answer score corresponding to the user answer data according to the preset answer score determination method;

    Determine a corresponding target comprehensive score according to the target answer score and the target facial expression score;

    Correspondingly update the existing total score of the answers according to the target comprehensive score; and

    When the updated total answer score meets the preset stop condition, the current question and answer process is stopped, and corresponding prompt information is sent to the terminal.
16. One or more non-volatile computer-readable storage media storing computer-readable instructions, which when executed by one or more processors, cause the one or more processors to perform the following steps:

    Receive user answer data and user answer images sent by the terminal;

    Determine whether the user's answer data is correct according to a preset determination method;

    When it is determined that the user answer data is correct, the target expression score corresponding to the user answer image is determined according to a preset expression score determination method;

    Determine the target question data according to the preset question data determination method corresponding to the target facial expression score; and

    Send the target question data to the terminal for display.
17. The storage medium according to claim 16, wherein the user answer image includes a user answer image; and when it is determined that the user answer data is correct, the answer to the user is determined according to a preset expression score determination method The target expression score corresponding to the image, including:

    When it is determined that the user answer data is correct, correspondingly obtain the user answer image according to the user answer image;

    Input the user answering image into a pre-trained micro-expression recognition model for prediction, and obtain corresponding user micro-expressions; and

    The user's micro-expression is input into a pre-trained expression score prediction model for prediction to obtain a target expression score.
18. The storage medium according to claim 16, wherein the user answering image includes a user answering video; and when it is determined that the user answering data is correct, the answering to the user is determined according to a preset expression score determination method The target expression score corresponding to the image, including:

    When it is determined that the user answer data is correct, correspondingly obtain the user answer video according to the user answer image;

    Extract a preset number of video frames from the user answer video according to a preset extraction method;

    Separately determine the expression score corresponding to each of the video frames; and

    A target expression score corresponding to the user answering image is determined according to each of the expression scores.
19. The storage medium according to claim 16, wherein the determining the target question data according to the preset question data determination method corresponding to the target expression score includes:

    Comparing the target expression score with a preset expression score; and

    When the target expression score reaches the preset expression score, the target question data of the preset question type is selected from the preset question bank.
20. The storage medium according to claim 16, wherein the user answer data and the user answer image correspond to current question data; and the target is determined according to the preset question data determination method corresponding to the target expression score Question data, including:

    Comparing the target expression score with a preset expression score; and

    When the target expression score is lower than the preset expression score, the target question data corresponding to the current question data is selected according to the preset selection method according to the current question data.

Images