CN115511556A

CN115511556A - Questionnaire processing method and device, electronic equipment and storage medium

Info

Publication number: CN115511556A
Application number: CN202211010067.4A
Authority: CN
Inventors: 陈丽霞; 刘勇成; 胡志鹏; 袁思思; 程龙
Original assignee: Netease Hangzhou Network Co Ltd
Current assignee: Netease Hangzhou Network Co Ltd
Priority date: 2022-08-22
Filing date: 2022-08-22
Publication date: 2022-12-23

Abstract

The application provides a questionnaire processing method and device, electronic equipment and a storage medium, and relates to the technical field of computers. According to the method, response results of research users for all questions in a questionnaire are obtained, game feature data of the research users corresponding to all the questions in the questionnaire are respectively input into a first model generated in advance, prediction results of all the questions in the questionnaire are obtained, and then whether the response results of the research users for all the questions are valid or not is determined according to the prediction results of all the questions in the questionnaire, so that valid response results in the questionnaire are reserved or invalid response results in the questionnaire are eliminated. According to the technical scheme, the complex process of manually picking and identifying the answered questionnaires is saved, and the recovery efficiency of the questionnaires is improved.

Description

Questionnaire processing method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for processing a questionnaire, an electronic device, and a storage medium.

Background

With the continuous development of internet technology, more and more users put forward higher requirements for services provided by providers when using the internet, and the providers ask for user opinions by establishing contact with the users in order to meet user experience and perfect service design, thereby achieving the purpose of continuously perfecting the services.

In the prior art, after a supplier makes a questionnaire, the questionnaire is delivered to a user terminal so that a user replies, after the reply is completed, the supplier terminal obtains the reply result, and relevant personnel screen the reply result to obtain the reply result of an available questionnaire.

However, the manual screening method is inefficient because of the large number of questions in the questionnaire or/and the large number of responses caused by the large number of users participating in the questionnaire.

Disclosure of Invention

In view of this, the present application provides a method, an apparatus, an electronic device, and a storage medium for processing a questionnaire, so as to overcome the problems that the number of questions in the questionnaire is large, or/and the number of users participating in the questionnaire is large, which results in a large number of response results, and the efficiency of a manual method for screening effective response questionnaires is low.

A first aspect of the embodiments of the present application provides a method for processing a questionnaire, where the method includes:

obtaining the response results of research users for each question in a questionnaire, wherein the questionnaire comprises at least one question;

inputting game feature data of research users corresponding to each topic of the questionnaire into a pre-generated first model respectively to obtain a prediction result of each topic in the questionnaire, wherein the first model is obtained by training according to the game feature data of at least one user corresponding to each topic;

and determining whether the response results of the investigation users to the various questions are valid according to the prediction results of the various questions in the questionnaire so as to retain the valid response results in the questionnaire or eliminate the invalid response results in the questionnaire.

A second aspect of the embodiments of the present application provides a processing apparatus for questionnaires, the processing apparatus including:

the system comprises an acquisition module, a search module and a search module, wherein the acquisition module is used for acquiring the response results of research users for each topic in a questionnaire, and the questionnaire comprises at least one topic;

the determining module is used for respectively inputting the game characteristic data of the research users corresponding to the questions of the questionnaire into a pre-generated first model so as to obtain the prediction results of the questions in the questionnaire, wherein the first model is obtained by respectively training according to the game characteristic data of at least one user corresponding to each question;

and the processing module is used for determining whether the response results of the investigation users for the various questions are valid according to the prediction results of the various questions in the questionnaire so as to retain the valid response results in the questionnaire or remove the invalid response results in the questionnaire.

A third aspect of the embodiments of the present application further provides an electronic device, including: a processor, a memory;

the memory stores computer-executable instructions;

the processor executes the computer-executable instructions to cause the electronic device to perform the method of processing questionnaires as described in the first aspect above.

The fourth aspect of the embodiments of the present application further provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the computer-readable storage medium is configured to implement the method for processing the questionnaire as described in the first aspect.

According to the technical scheme provided by the embodiment of the application, the response results of research users for all questions in the questionnaire are obtained, the questionnaire comprises at least one question, the game characteristic data of the research users corresponding to all the questions in the questionnaire are respectively input into a pre-generated first model to obtain the prediction results of all the questions in the questionnaire, the first model is obtained by training according to the game characteristic data of at least one user corresponding to all the questions, and then whether the response results of the research users for all the questions are valid is determined according to the prediction results of all the questions in the questionnaire, so that the valid response results in the questionnaire are reserved or the invalid response results in the questionnaire are eliminated. According to the technical scheme, a model for training and predicting possible results of the questionnaire is set out according to game feature data of the user corresponding to the question, so that judgment of response results meeting expectations is achieved, the situation that the user fills in the questionnaire maliciously or randomly is avoided, the tedious picking and identifying process of manually answering the questionnaire is further saved, and the recovery efficiency of the questionnaire is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments of the present application will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive exercise.

Fig. 1 is a schematic view of an application scenario of processing a questionnaire provided in an embodiment of the present application;

fig. 2 is a schematic flowchart of a first embodiment of a method for processing a questionnaire provided in an embodiment of the present application;

fig. 3 is a schematic flowchart of a second method for processing a questionnaire provided in an embodiment of the present application;

fig. 4 is a schematic flowchart of a third embodiment of a method for processing a questionnaire provided in the embodiment of the present application;

fig. 5 is a schematic flowchart of a fourth embodiment of a method for processing a questionnaire provided in the embodiment of the present application;

fig. 6 is a schematic flowchart of a fifth embodiment of a processing method of a questionnaire provided in an embodiment of the present application;

fig. 7 is a schematic flowchart of a sixth embodiment of a processing method of a questionnaire provided in an embodiment of the present application;

fig. 8 is a schematic flowchart of a seventh embodiment of a processing method of a questionnaire provided in an embodiment of the present application;

FIG. 9 is a schematic diagram of a device for processing questionnaires provided in an embodiment of the present application;

fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

In order to enable those skilled in the art to better understand the technical solution of the present application, the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. This application is capable of embodiments in many different forms than those described above and it is therefore intended that all such other embodiments, which would be within the scope of the present application and which are obtained by a person of ordinary skill in the art based on the embodiments provided herein without the exercise of inventive faculty, be covered by the present application.

It should be noted that the terms "first," "second," "third," and the like in the claims, the description, and the drawings of the present application are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. The data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Before introducing the embodiments of the present application, the terms and background related to the present application will be explained as follows:

user portrait: an effective tool for sketching appeal and design direction of target users and contact users. The user portrait abstracts each concrete information of the user into tags, and concreties the user image by using the tags, thereby providing targeted services for the user.

Common user portrayal modes include: group quantitative classification, cluster analysis, qualitative interview, and the like.

With the continuous development of internet technology, when more and more users use the internet, the requirements for services provided by providers are continuously increased, and in order to improve user experience and perfect service design, the providers establish contact with the users to ask for user opinions, and the providers continuously perfect the services.

Taking game service as an example, in order to perfect game design and improve game experience of players, it is very important to establish direct or indirect communication between game manufacturers and players to obtain the views of players on games. Currently, game vendors often choose to obtain information about players in a gaming system by issuing special questionnaires designed by user researchers to the players.

The prior art scheme mainly has the following defects:

1) Although there are trigger conditions in the questionnaire, some questions need to be triggered by a question-specific pre-choice (e.g., a question in the questionnaire is set as "you buy skin often", when the user selects "yes", it is triggered "whether skin you buy often is story-level skin");

this triggering method can generate questionnaires with diversity limited by the number of combinations of triggering conditions, which is not flexible and diverse enough, and loses the pertinence of questionnaires.

2) After the questionnaire is recovered by the system, technicians need to spend extra time to manually remove invalid problems or questionnaires which affect the investigation result, and the system is time-consuming and labor-consuming;

3) When the number of invalid questionnaires is too large, questionnaire investigation needs to be arranged (at least once) again, and the cost for acquiring valid information is increased.

In view of the above technical problems, the inventors of the present invention have found that: in the prior art, the issuing of the questionnaire is not targeted, so that the validity of the questionnaire is reduced, and if the questionnaire can be issued, the label is labeled, and the questionnaire is issued to a user corresponding to the label, so that the validity of the questionnaire can be improved. Furthermore, in order to prevent the questionnaire responses of the users from obtaining some rewards only for the process, the feature data of one or more users corresponding to the tags to which the users belong can be obtained in a big data mode, and a model capable of judging the validity of the questionnaire responses of the users is generated to realize the judgment of the questionnaire validity, so that only valid questionnaires are reserved, and the situation of heavy workload when invalid questionnaires are manually selected is avoided.

Based on the problems in the prior art, fig. 1 is a schematic view of an application scenario of processing a questionnaire provided in an embodiment of the present application, so as to solve the technical problems. As shown in fig. 1, the application scenario diagram includes: a terminal device 11 and an electronic device 12.

The number of the terminal devices 11 may be one or more, that is, the number of the user accounts may be one or more; the terminal device 11 may be a mobile phone, a computer, a tablet, a smart watch, or other devices with a display function.

Taking a game service as an example, the electronic device 12 may be a network-side device that provides a service for a certain game application installed on the terminal device 11.

In one possible implementation, the technician tags the frequent purchases to improve the experience of the gaming application to the user, such as a shopping recommendation experience. The electronic device 11 can read the player tag in the user image database, and use the player with the matched shopping frequency as the object of issuing the questionnaire corresponding to the shopping recommendation, that is, the research user.

Further, the terminal corresponding to the research user may be the terminal device 11, the electronic device 12 sends the questionnaire to the terminal device 11, and after the research user replies, the terminal device 11 returns the reply result to the electronic device 12.

The electronic device 12 records the prediction result of the questionnaire in advance, and compares the prediction result with the response result to judge the validity of the questionnaire.

It should be understood that: the above application scenarios are only examples and are not limiting to the present application, and moreover, the content not disclosed in the application scenarios is referred to the following embodiments.

The technical solution of the present application will be described in detail by specific examples. It should be noted that the following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

Fig. 2 is a schematic flow chart of a first method for processing a questionnaire provided in the embodiment of the present application. As shown in fig. 2, the method for processing the questionnaire may include the following steps:

and step 21, obtaining the response results of the research users for each question in the questionnaire.

In this step, the validity of the response result corresponding to the questionnaire filled by the research user needs to be judged, and first, each question in the questionnaire needs to be acquired.

Each questionnaire comprises at least one question, and the answer result of each question is the question answer obtained by the research user filling in the questionnaire, can be the option of the research user in the questionnaire when answering the question, can be single-choice, multi-choice, or the answer text of an open question.

And step 22, respectively inputting the characteristic data of the research user corresponding to each question of the questionnaire into a pre-generated first model so as to obtain the prediction result of each question in the questionnaire.

The first model is obtained by training according to game feature data of at least one user corresponding to each topic.

In this step, in order to compare the prediction result with the response result, feature data of the research user corresponding to the topic corresponding to the response result needs to be input into the first model generated in advance, so as to obtain the prediction result of the topic.

Optionally, in implementation, for the same topic (problem), the required model inputs game feature data of at least one user corresponding to the topic, and outputs an option of the topic to construct a first model, and when the model is used, the game feature data of the research user is acquired, and the game data feature of the research user is input to the first model to obtain a possible option of the topic, that is, a prediction result.

In one possible implementation, taking the theme of satisfaction of the merchandise as an example, the game feature data of the research user can be the data of participation/purchase, the first participation interval, the use/participation frequency, the latest participation/use, the last use/participation time from now and the like.

Optionally, a machine learning algorithm, e.g., a classification algorithm, may be employed in training the first model.

In one possible implementation, taking the prediction of the answer to the question whether a new event or good is satisfied as an example:

the predicted results of the topics can be sorted and expressed as: is very satisfactory (60%), generally satisfactory (20%), generally (10%), not satisfactory (10%).

In another possible implementation, take the example of suggesting a class openness problem (e.g., please suggest to the campaign):

the predicted results of the topics can be sorted and expressed as: positive/negative, or positive/negative/neutral.

Example 1, positive may denote: there are suggestions; can be represented by: there is no suggestion; negative can mean: the user is confused to answer.

Example 2, positive may denote: positive advice; can be represented by: a Chinese suggestion; negative can mean that: and (4) negative suggestion.

And step 23, determining whether the response results of the investigation users for the various questions are valid according to the prediction results of the various questions in the questionnaire so as to retain the valid response results in the questionnaire or remove the invalid response results in the questionnaire.

In this step, after the electronic device receives the reply result, it needs to determine whether the reply result is valid, that is, whether the reply result is a result after the investigation user answers seriously, and then the electronic device performs a retention process or a rejection process on the reply result.

Optionally, the answer result of each topic by the research user is compared with the predicted result of each topic in the questionnaire corresponding to the research user, so as to determine whether the answer result of each topic is valid.

In one possible implementation, if the response result meets the predetermined expectation of the predicted result, the response result is determined to be valid, and the valid response result is retained.

In another possible implementation, if the reply result does not meet the preset expectation of the prediction result, a next determination needs to be performed to determine to retain or reject the reply result, which is provided by the following embodiments and will not be described herein again.

Further, a questionnaire containing no invalid response results is retained, or a questionnaire containing invalid response results is rejected.

According to the technical scheme provided by the embodiment of the application, the method includes the steps that answer results of research users for all questions in a questionnaire are obtained, the questionnaire comprises at least one question, game feature data of the research users corresponding to the questions of the questionnaire are respectively input into a first model generated in advance to obtain predicted results of the questions in the questionnaire, the first model is obtained through training according to the game feature data of the users corresponding to the questions, and then whether the answer results of the research users for the questions are valid or not is determined according to the predicted results of the questions in the questionnaire, so that valid answer results in the questionnaire are reserved or invalid answer results in the questionnaire are eliminated. In the technical scheme, a model for training and predicting possible results of the questionnaire is set out by using game feature data of the user corresponding to the question, so that judgment of answer results meeting expectations is realized, the situation that the user maliciously or randomly fills in the questionnaire is avoided, further, the tedious picking and identifying process of manually answering the questionnaire is saved, and the recovery efficiency of the questionnaire is improved.

Based on the above embodiments, fig. 3 is a schematic flow chart of a second embodiment of a processing method of a questionnaire provided in the embodiment of the present application. As shown in fig. 3, before step 21, the method for processing a questionnaire may further include the steps of:

step 31, determining a user label of the research user based on the user portrait.

In this step, in order to increase the effectiveness and pertinence of the questionnaire, it is first necessary to determine the user tags of the survey users, that is, to which kind of users the questionnaire needs to be delivered, so as to obtain feedback.

Optionally, the user portrait in the user portrait database may be drawn by using a clustering analysis in combination with technical means such as qualitative interview and quantitative data analysis, that is, a user label corresponding to the user is marked.

For example, table 1 is a user tag schematic table of a research user provided in the embodiment of the present application, as shown in table 1:

table 1:

and step 32, extracting questions corresponding to the questionnaire labels with consistent user labels from the question bank to generate the questionnaire.

In this step, a plurality of questions are stored in the question bank in advance, each question has a corresponding questionnaire tag, and then questions corresponding to questionnaire tags with consistent user tags are extracted from the question bank of the questionnaire to generate a questionnaire.

Optionally, as an example, table 2 shows topics corresponding to questionnaire tags provided in the embodiment of the present application, as shown in table 2:

table 2:

questionnaire label	Question number
		Baoma	1、3、5、8、9、11、……
Easily purchased men's clothing	2、3、6、9、16、……
		Easy-to-purchase electricity device	1、6、9、8、7、2、3
Wine easy to buy	3、5、6、12、25、45
		Active players	48、69、58、44、96
New registered user	5、9、11、16、17、……

Further, the question corresponding to the questionnaire tag that matches the user tag is extracted, and taking user 1 (research user) in table 1 as an example, the extracted questionnaire tag is: precious mother, easily purchase men's clothing and easily purchase electric appliance, the corresponding title serial number is: 1. 3, 5, 8, 9, 11, 2, 6, 16, 7, … ….

Alternatively, the extraction rule may be random extraction, extraction according to odd numbers, or extraction of an unusual topic, and the number of extractions may be set at will.

Specifically, the questions may be 9, 11, or 20 questions numbered first, or any of the odd-numbered 9, 12, or 20 questions, and the questionnaire may be generated from the extracted questions.

Each label can correspond to a plurality of topics, one topic can also correspond to a plurality of labels, and the corresponding relationship is not limited.

Step 33, pushing the questionnaire to the investigation user.

In this step, after the questionnaire is generated in the above step, the questionnaire is pushed to the investigation user.

Optionally, in the implementation process of the whole solution, the number of research users may be one or more (that is, questionnaires are issued to multiple users), and the illustration push is performed by taking 3 research users as an example. Table 3 is an example of questionnaire push, as shown in table 3:

table 3:

furthermore, after the electronic device side generates the questionnaire, the questionnaire is pushed to the investigation user, and after the investigation user selects to fill in the questionnaire, the questionnaire is arranged on the graphical user interface of the investigation user's terminal device according to a certain rule, and then the investigation user answers.

In one possible implementation, each page of the graphical user interface shows 10 topics, and when more than 10 topics are in the questionnaire, a plurality of pages are shown, with no more than 10 being shown in a page.

According to the technical scheme, the user tags of the research users are determined based on the user portrait, questions corresponding to the questionnaire tags with the consistent user tags are extracted from the question bank, the questionnaire is generated, and the questionnaire is pushed to the research users. According to the technical scheme, the questions corresponding to the questionnaire labels which are consistent with the user labels of the investigation users are pushed to the investigation users, so that the pertinence of questionnaire putting is improved, and the reliability of response results of questionnaires is improved.

The construction and application of the first model are briefly described below:

fig. 4 is a schematic flow chart of a third embodiment of a processing method of a questionnaire provided in the embodiment of the present application. As shown in fig. 4, the processing of the questionnaire further includes the steps of:

and step 41, obtaining game characteristic data of at least one user corresponding to the theme.

In this step, the questionnaire labels corresponding to the questions are obtained from the game data, the game log and other information, and the game feature data of at least one user consistent with the questionnaire labels is extracted, so that training data is provided for the subsequent construction of the first model.

Optionally, as an example, table 4 is an example of game feature data that needs to be acquired corresponding to different titles, as shown in table 4:

table 4:

topic of questions	Game feature data of at least one user
		Whether the goods are satisfactory	Number of purchases per quarter … …
Whether new activity is satisfied	Interval time length … … for first participation in first-onset activity
		Whether or not to like to purchase electronic products	Number of purchases per quarter, whether the last 30 days purchased … …
Whether a player is active	Number of logins and traffic utilization … … in last 30 days

data of whether to participate or purchase, whether to participate for the first time, the first time interval, the use or participation frequency, whether to participate or use recently, the last time or the participation time, and the like are extracted as game characteristic data.

and extracting game data such as the participation frequency of the players, the winning proportion of the players, whether the players participate recently, the losing time, whether the players return back and the like as game characteristic data.

And 42, training a machine learning algorithm model according to the game characteristic data of at least one user to obtain a first model.

In this step, the first model is a specified model of a topic, and functions as: the game characteristic data of a plurality of users are used for classifying the output result of the title, and when the game characteristic data of a certain user is used in the subsequent process, the specified prediction result can be obtained.

Optionally, as an example, table 5 is an example of an output result of the game feature data after training, as shown in table 5 (the title is how often bottled water is purchased):

table 5:

wherein, the trained features can be the judgment standard of the judged classification result.

Optionally, in the generation process of the first model, the game feature data of at least one user may be used, the classification algorithm is used to classify the game feature data, each classification interval corresponds to a different classification result, and after the game feature data of a certain user is subsequently input, a classification result, that is, a prediction result, may be obtained.

In one possible implementation, the game log may provide rich player behavior data, the big data engineer may select different data and algorithms to try, train to obtain different models, and then evaluate the models to obtain the most accurate predicted model, i.e., the first model.

Specifically, behavior data is used as game feature data, a training set and a testing set are constructed, different algorithms are used for training to obtain a model- > testing the effect of the testing set- > adjusting the feature- > obtaining the model- > testing the effect- > repeating until a better model is obtained.

And step 43, inputting the game characteristic data of the research user corresponding to the theme into the first model to obtain the prediction result of the theme.

In this step, in order to compare the predicted result with the response result, the game feature data of the research user corresponding to the topic needs to be input into the first model, and the trained features can be compared with the game feature data to obtain the predicted result of the topic.

Optionally, as an example, table 6 is an example of game feature data of users and corresponding prediction results, as shown in table 6 (it should be understood that users in table 6 are users who make questionnaires, and some prediction results for different topics):

table 6:

it should be understood that: for one topic, the first model is obtained by training at least one game feature data of a user corresponding to the topic, and the model can judge what data features obtain what possible options.

Such as "which piece do you like the best fashionable dress? "what the first model actually needs to do is to use the behavior data of the current player to guess the favorite fashion identifier of the current player.

Thus, training with at least one user's data yields a first model that yields a specified result upon entry of a specified feature. The process of predicting the result is the answering process corresponding to the found question, the data of the research user who makes the questionnaire is extracted, the characteristic data required by the model is determined, and the characteristic data of the research user is input into the model, so that the possible selection (namely the predicted result) of the research player in the question is obtained.

In one possible implementation, the above prediction of the answer to the question of whether a new event or good is satisfied is taken as an example:

the predicted results of the topics can be sorted and expressed as: very satisfactory (60%), generally satisfactory (20%), generally (10%), unsatisfactory (10%).

In another possible implementation, the above-mentioned suggestion-like openness problem (e.g., please suggest to the campaign) is used as an example:

the predicted results of the topics can be sorted and expressed as: two positive/negative classes, or three positive/negative/neutral classes.

Example 1, positive may denote: there are suggestions; can represent that: there is no suggestion; negative can mean: the user can do a mess to answer.

Example 2, positive may denote: positive advice; can represent that: a Chinese suggestion; negative can mean: and (4) negative suggestion.

According to the technical scheme provided by the embodiment of the application, the game characteristic data of at least one user corresponding to a theme is obtained, the machine learning algorithm model is trained according to the game characteristic data of the at least one user to obtain the first model, and then the game characteristic data of the research user corresponding to the theme is input into the first model to obtain the prediction result of the theme. In the technical scheme, the model is trained by utilizing the game characteristic data of a plurality of users, so that the accuracy of the model in the subsequent prediction by utilizing the game characteristic data of the investigated users is increased, and a basis is provided for the judgment of the subsequent effective response result.

On the basis of the above example, the questionnaire may contain one or more types of topics as follows:

1) And selecting the single;

specifically, the option of one topic may be one.

2) Selecting more;

specifically, the options for one topic may be at least two.

3) And an open type.

Specifically, the reply content of a topic is supplemented by the user himself, without any option.

Further, fig. 5 is a schematic flow chart of a fourth embodiment of a method for processing a questionnaire provided in the embodiment of the present application. As shown in fig. 5, it is explained that, in the above step 23, if the answer result matches the preset expectation of the prediction result, the answer result is determined to be valid, and the valid answer result is retained.

In this step, since the predicted result is obtained by using the model trained from the game feature data of the user printed with the user tag, the behavior of the user corresponding to the user tag can be reflected to a greater extent, which can improve the validity of the answer result.

Alternatively, the preset expectation may be an option with a higher probability in the prediction result.

In one possible implementation, the prediction result may be 70% (key sensitivity set is low), 20% (key sensitivity set is moderate), 5% (key sensitivity set is high), and the preset expectation of the prediction result is the first two items, i.e., the answer result is that the key sensitivity set is low or the key sensitivity set is moderate, the answer result of the questionnaire is considered to be valid; if the answer result is the latter items, it needs to be further determined whether the answer result is available, which is given in the following embodiments.

The following possible implementations have no tandem order, and the triggers may be different for different types of topics:

1, the type of the answer result of the topic is single selection:

illustratively, a comparison is made as to whether the response hits in the top set of probabilities in the predicted result (i.e., X options are preset that are expected to be top in probabilities, where X is a positive integer greater than or equal to 1).

Optionally, as an example, table 7 is a first example of the ordering of the prediction results obtained after different topics are input into the first model and the corresponding preset expectations, as shown in table 7 (taking X as the first 2 options as an example):

table 7:

that is, when any of the first 2 options of the predicted result is identical to the answer result at the time of 70% (very satisfactory), 20% (satisfactory), 5% (normal), 5% (unsatisfactory), the answer result of the questionnaire is described to be valid and the valid answer result is retained.

And 2, the type of the answer result of the topic is multiple choices:

illustratively, a comparison is made as to whether the answer result misses the later set of probabilities in the predicted result (i.e., the preset expectation does not include the later Y options, where Y is a positive integer greater than or equal to 1).

Optionally, as an example, table 8 is a second example of the ordering of the prediction results obtained after different topics are input into the first model and the corresponding preset expectations, as shown in table 8 (taking Y as the last 1 option as an example):

table 8:

that is, the predicted results are 70% (good), 15% (good), 10% (recommended to others), and 5% (bad), and when the last Y (e.g., 1) options of the predicted results do not coincide with any of the options in the response results, the response results of the questionnaire are described as valid, and the valid response results are retained.

And 3, the type of the answer result of the topic is an open type:

illustratively, a sentence segmentation (e.g., jieba segmentation) is used to segment a reply result to obtain a segmentation result, a preset customized deactivation vocabulary is used to remove stop words in the segmentation result to obtain a target result, and then the target result is input to a model (i.e., a second model) constructed by a machine learning algorithm (e.g., an open source code version control System (SVN)) to obtain an answer emotion of a user.

Optionally, the target result is input to a second model to obtain an answer emotion, the second model is used for determining the emotion expressed by the research user in the open-type question, and if the answer emotion is consistent with the expression emotion in the prediction result, the answer result is determined to be valid.

Further, whether the answer emotion is consistent with the preset expectation of the prediction result is compared.

In one possible implementation, the predicted result may be a positive emotion, and the answer emotion obtained by processing the answer result is also positive, which indicates that the answer result of the questionnaire is valid.

According to the technical scheme provided by the embodiment of the application, the question types in the questionnaire are divided into single-selection, multi-selection and/or open types, so that different judgment modes are set for different question types, and the judgment on whether the answer results of research users are effective answer results is more accurately realized.

Based on the above embodiments, fig. 6 is a schematic flow chart of a fifth embodiment of a processing method of a questionnaire provided in the embodiment of the present application. As shown in fig. 6, how to determine that the response result of the questionnaire is valid if the response result does not meet the predetermined expectation of the predicted result in step 23 is described.

Optionally, for a case that the reply result does not meet the preset expectation of the prediction result, the following steps may be performed:

and step 61, obtaining the number of times that the question corresponding to the response result has been answered in the questionnaire.

In this step, in order to avoid the investigation user's random answer (for example, randomly filling in to obtain the reward of answer issued by the system), when it is determined that the answer result does not meet the preset expectation of the predicted result, the question may be added to the questionnaire again, so that the investigation user may reply to the question again to consider whether the investigation user randomly answers the question.

At this time, the number of times that the question corresponding to the preset expected response result which does not accord with the prediction result has been answered is obtained.

In one possible implementation, the number of times that the topic corresponding to the obtained response result has been answered in the questionnaire is 1.

And step 62, if the times are less than or equal to the times threshold, readjusting the order of the options in the question corresponding to the reply result which does not accord with the preset expectation of the prediction result.

In this step, in order to avoid the fluctuation of the mood of the user caused by the excessive occurrence of the same topic in the questionnaire, a threshold value of the occurrence is set at this time, and the occurrence of the same topic in the questionnaire is limited.

Further, when the number of times is smaller than or equal to the number threshold, in order to avoid the investigation user filling in at will and increase the experience of the investigation user, the order of the options in the title is readjusted.

Optionally, as an example, table 9 is an illustration of the number of times that the topic has appeared and a threshold number of times, as shown in table 9:

table 9:

question number	Number of times of answer	Number of times threshold	Whether to adjust the order of the options
				Z	1	3	Is that
Q	4	3	Whether or not

That is, the threshold of the number of times may be 3, and following the above steps, for example, where 1 is smaller than 3, the order of the options in the topic corresponding to the reply result is adjusted.

For example, the content of the item corresponding to the theme is "satisfactory, general, and unsatisfactory", and in this case, the content is adjusted to "unsatisfactory, satisfactory, general".

And step 63, supplementing the topics with the adjusted option sequence to a to-be-released list in the questionnaire so that the research user can reply again.

In this step, the questions with the adjusted selection order are supplemented to the to-be-placed list in the questionnaire, so that the questions are placed to the research user again, and the research user answers again, so as to facilitate effective judgment of the response result of the questions.

According to the technical scheme provided by the embodiment of the application, the times that the questions corresponding to the response results are answered in the questionnaire are obtained, if the times are smaller than or equal to the time threshold value, the option sequence in the questions corresponding to the response results which do not accord with the preset expectation of the prediction results is readjusted, and the questions with the adjusted option sequence are supplemented to the to-be-released list in the questionnaire, so that the investigation user can answer again. In the technical scheme, the option sequence of the questions corresponding to the reply result which is uncertain whether the reply result is valid is adjusted and released to the user again, so that reference can be provided for more accurately determining whether the reply result is valid in the follow-up process.

On the basis of the foregoing embodiment, fig. 7 is a flowchart illustrating a sixth embodiment of a method for processing a questionnaire according to the embodiment of the present application. As shown in fig. 7, after step 61, the method for processing a questionnaire may further include the steps of:

wherein, the embodiment is as follows: and under the condition that the times are greater than the time threshold value, determining whether the response results are valid according to all response results of the questions corresponding to the preset expected response results which do not accord with the prediction results so as to retain the valid response results in the questionnaire or remove the invalid response results in the questionnaire.

And step 71, when the times are greater than the time threshold, respectively converting the options of the questions corresponding to the preset expected reply results which do not accord with the prediction results into the equal interval sequence numerical representation to obtain the numbers corresponding to different options.

In this step, in order to ensure whether the response result of the same question is reliable, the options corresponding to the question are respectively converted into the equal interval sequence numerical representation, and the numbers corresponding to different options are obtained.

In one possible implementation, the option for topic Z can be U, V, W.

In one possible implementation, the choice for topic Q can be U, V, W, E.

Optionally, as an example, table 10 is an example of all the response results corresponding to different topics, as shown in table 10:

table 10:

question number	Options for the subject	Converted number
			Z	U、V、W	1、2、3
Q	U、V、W、E	1、2、3、4

That is, for the topic Z, the option U corresponds to 1, the option V corresponds to 2, and the option W corresponds to 3; for the title Q, the option U corresponds to 1, the option V corresponds to 2, the option W corresponds to 3, and the option E corresponds to 4.

And step 72, determining the variance of the numbers corresponding to the options indicated by all the response results according to the numbers corresponding to the options indicated by all the response results.

In this step, for a topic, when the number of times that the topic has been answered is greater than the threshold number of times, the numbers corresponding to the options indicated by the answer results of the answered topic are counted, and the variance of the numbers is obtained.

The variance is used to measure the deviation degree between each digit and the mean value of all digits, i.e. the variation degree of multiple response results representing the same question.

In one possible implementation, the 3 answered options for topic Z correspond to numbers 1, 2, and 3, respectively, with a variance of about 0.67.

In one possible implementation, the numbers corresponding to the 3 answered options of the question Q are 4, 4 or 1, respectively, and the variance is 0.

And step 73, determining whether the response result is valid according to the variance and the tolerance threshold value so as to retain the valid response result in the questionnaire or remove the invalid response result in the questionnaire.

In this step, when the deviation degree is greater than the tolerance threshold, all the reply results which do not meet the preset expectation of the predicted result are invalid; when the deviation degree is not larger than the tolerance threshold, all the reply results which do not meet the preset expectation of the predicted result are valid.

In one possible implementation, the tolerance threshold is set to 0.1, the variance of the numbers corresponding to the options indicated by all the response results of topic Q is 0, and 0 is less than 0.1, then the research user expresses a real idea, which is marked as a valid response result, to retain the valid response result.

In one possible implementation, the tolerance threshold is set to 0.1, the variance of all responses for topic Z is 0.67, and 0.67 is greater than 0.1, then the investigator does not express a true idea, marks it as an invalid response, and rejects it.

According to the technical scheme provided by the embodiment of the application, when the times are greater than the times threshold value, the options of the items corresponding to the preset expected response results which do not accord with the prediction results are respectively converted into the numbers corresponding to the equal-interval sequences to obtain the numbers corresponding to different options, the variance of the numbers corresponding to the options indicated by all the response results is determined according to the numbers corresponding to the options indicated by all the response results, and whether the response results are effective or not is determined according to the variance and the tolerance threshold value so as to retain the effective response results in the questionnaire or eliminate the invalid response results in the questionnaire. According to the technical scheme, the real ideas of the user for the same question are judged by using the change degrees among all the response results of the same question, whether the response of the question is effective or not can be more accurately determined, automatic elimination of dirty data influencing the investigation result is completed, the effectiveness of the questionnaire is effectively improved, and the work efficiency of questionnaire recycling is greatly improved.

On the basis of the foregoing embodiment, fig. 8 is a flowchart illustrating a seventh embodiment of a processing method for a questionnaire provided in the embodiment of the present application. As shown in fig. 8, the processing method of the questionnaire includes the following steps:

step 1, starting;

step 2, extracting a user tag from a user image database;

step 3, judging whether the user label is consistent with the questionnaire label, if so, executing the step 4; if not, executing the step 17;

step 4, obtaining a model corresponding to the label;

step 5, extracting game characteristic data in a time interval which meets the requirement of a research user recently for picking from a database;

step 6, inputting the game characteristic data into a model to obtain a prediction result;

wherein the first model is trained based on game feature data of at least one player.

Step 7, putting the questions to the investigation user by using a certain rule;

step 8, waiting for completion of the questionnaire (whether completion is achieved and page turning is clicked);

step 9, comparing whether the prediction result and the reply result have deviation or not, if yes, executing step 10; if not, executing the step 12;

step 10, judging whether the number of times of the occurrence of the question does not exceed a number threshold, if so, executing step 11; if not, executing the step 12;

step 11, changing the option times of the question, and putting the question again;

step 12, after the execution of the process is finished, recovering all questions and corresponding reply results;

step 13, comparing the response results with the deviation, and determining the variance of the numbers corresponding to the response results with the deviation;

step 14, judging whether the variance is smaller than a tolerance threshold value, if so, executing step 15; if not, executing the step 16;

step 15, marking as a valid answer result, and executing step 17;

step 16, marking as an invalid response result;

and step 17, ending.

The technical scheme provided by the embodiment of the application has the beneficial effects that the beneficial effects are seen in the embodiment.

On the basis of the above method embodiment, fig. 9 is a schematic diagram of a processing apparatus for questionnaires provided in an embodiment of the present application, and as shown in fig. 9, the processing apparatus for questionnaires includes: an acquisition module 91, a determination module 92 and a processing module 93.

The obtaining module 91 is configured to obtain a response result of the research user for each topic in a questionnaire, where the questionnaire includes at least one topic;

the determining module 92 is configured to input the game feature data of the research users corresponding to the respective topics of the questionnaire into a pre-generated first model respectively, so as to obtain a prediction result of each topic in the questionnaire, where the first model is obtained by training the game feature data of at least one user corresponding to each topic;

and the processing module 93 is configured to determine whether the response result of the research user for each question is valid according to the prediction result of each question in the questionnaire, so as to retain the valid response result in the questionnaire or remove the invalid response result in the questionnaire.

In one possible design of the embodiment of the present application, the processing module 93 is specifically configured to:

if the reply result meets the preset expectation of the prediction result, determining that the reply result is valid, and keeping the valid reply result;

or the like, or, alternatively,

if the answer result does not accord with the preset expectation of the prediction result, acquiring the number of times that the question corresponding to the answer result has been answered in the questionnaire;

and if the frequency is greater than the frequency threshold value, determining whether the response result is valid according to all response results of the questions corresponding to the preset expected response results which do not accord with the prediction result so as to retain the valid response results in the questionnaire or remove the invalid response results in the questionnaire.

In this possible design, the processing module 93 determines whether the response result is valid according to all the response results of the questions corresponding to the response results that do not meet the preset expectation of the prediction result, so as to retain valid response results in the questionnaire or remove invalid response results in the questionnaire, and is specifically configured to:

respectively converting options of the questions corresponding to the preset expected response results which do not accord with the prediction results into equal interval sequence numerical representation to obtain numbers corresponding to different options;

determining the variance of the numbers corresponding to the options indicated by all the response results according to the numbers corresponding to the options indicated by all the response results;

and determining whether the response result is valid according to the variance and the tolerance threshold value so as to retain the valid response result in the questionnaire or reject the invalid response result in the questionnaire.

Optionally, the processing module 93, according to the variance and the tolerance threshold, determines whether the response result is valid, so as to retain a valid response result in the questionnaire or remove an invalid response result in the questionnaire, and is specifically configured to:

if the variance is smaller than the tolerance threshold, determining that all preset expected reply results which do not accord with the prediction result are valid, and keeping valid reply results;

and if the variance is greater than or equal to the tolerance threshold, determining that all preset expected reply results which do not accord with the predicted result are invalid, and removing the invalid reply results.

Optionally, the processing module 93 is further configured to:

if the times are less than or equal to the times threshold value, readjusting the order of options in the questions corresponding to the preset expected reply results which do not accord with the predicted results;

and supplementing the topics with the adjusted option sequence to a to-be-released list in a questionnaire so that the research user answers again.

In another possible design of the embodiment of the present application, the answer result of the question in the questionnaire is a single-choice type, and the preset expectation is: hit the first X sets of the probability magnitude ordering in the prediction result, X being an integer greater than 0;

the processing module 93 determines that the answer result of the questionnaire is valid when the answer result meets the preset expectation of the prediction result, and retains the valid answer result, which is specifically used for:

and if the answer result hits the first X sets with the probability size in the predicted result, determining that the answer result is valid and reserving the valid answer result.

In another possible design of the embodiment of the present application, the answer result of the question in the questionnaire is a type of multiple choices, and the preset expectation is: the last Y sets of the probability size ordering in the miss prediction result are obtained, and Y is an integer larger than 0;

and if the answer result does not hit the last Y sets in the probability size sequence in the prediction result, determining that the answer result is valid and reserving the valid answer result.

In yet another possible design of the embodiment of the present application, the answer result of the question in the questionnaire is an open type, and the preset expectation is: the emotion is consistent with the expression emotion in the prediction result;

segmenting words of the answer result of the question by using sentence segmentation to obtain a segmentation result;

removing stop words in the word segmentation result by using a preset user-defined stop word list to obtain a target result;

inputting the target result into a second model to obtain answer emotion, wherein the second model is used for determining emotion expressed by a research user in an open question;

and if the answer emotion is consistent with the expression emotion in the prediction result, determining that the response result is valid, and keeping the valid response result.

In yet another possible design of the embodiment of the present application, the processing module 93 is further configured to:

determining a user label of a research user based on the user representation;

extracting questions corresponding to the questionnaire labels with consistent user labels from a question bank to generate questionnaires;

and the pushing module is used for pushing the questionnaire to the investigation user.

In another possible design of the embodiment of the present application, the obtaining module 91 is further configured to obtain game feature data of at least one user corresponding to a topic;

the processing module 93 is further configured to train a machine learning algorithm model according to the game feature data of at least one user to obtain a first model.

The processing device for questionnaires provided in the embodiment of the present application can be used to execute the technical solutions corresponding to the processing methods for questionnaires in the above embodiments, and the implementation principles and technical effects are similar, and are not described herein again.

It should be noted that the division of the modules of the above apparatus is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these modules can be realized in the form of software called by processing element; or may be implemented entirely in hardware; and part of the modules can be realized in the form of calling software by the processing element, and part of the modules can be realized in the form of hardware. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.

Fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 10, the electronic device may include: a processor 101, a memory 102, and computer program instructions stored on the memory 102 and operable on the processor 101.

The processor 101 executes computer-executable instructions stored by the memory 102, causing the processor 101 to perform the aspects of the embodiments described above. The processor 101 may be a general-purpose processor including a central processing unit CPU, a Network Processor (NP), and the like; but also a digital signal processor DSP, an application specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components.

The memory 102 is connected to the processor 101 via a system bus and communicates with each other, and the memory 102 is used for storing computer program instructions.

Optionally, the structure of the electronic device further includes: and a transceiver 103, wherein the transceiver 103 is connected with the processor 101 through a system bus and completes communication with each other.

In implementation, the transceiver 103 may correspond to the obtaining module 91 and the pushing module in the embodiment shown in fig. 9.

The system bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.

The electronic device provided in the embodiment of the present application may be configured to execute the technical solution corresponding to the processing method of the questionnaire in the above embodiment, and the implementation principle and the technical effect are similar, which are not described herein again.

The embodiment of the application further provides a chip for running the instructions, and the chip is used for executing the technical scheme of the processing method for the questionnaire in the embodiment.

An embodiment of the present application further provides a computer-readable storage medium, where a computer instruction is stored in the computer-readable storage medium, and when the computer instruction runs on an electronic device, the electronic device is enabled to execute the technical solution of the processing method for questionnaires in the foregoing embodiments.

The computer-readable storage medium described above may be implemented by any type of volatile or non-volatile storage device or combination thereof, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk. Readable storage media can be any available media that can be accessed by a general purpose or special purpose electronic device.

It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Although the present application has been described with reference to the preferred embodiments, it is not intended to limit the present application, and those skilled in the art can make variations and modifications without departing from the spirit and scope of the present application, therefore, the scope of the present application should be determined by the claims that follow.

Claims

1. A method of processing a questionnaire, comprising:

2. The method according to claim 1, wherein the determining whether the answer result of the survey user for each topic is valid according to the predicted result of each topic in the questionnaire, so as to retain the valid answer result in the questionnaire or eliminate the invalid answer result in the questionnaire comprises:

or the like, or a combination thereof,

and if the times are larger than a time threshold value, determining whether the response results are valid according to all response results of the questions corresponding to the preset expected response results which do not accord with the prediction results, so as to retain valid response results in the questionnaire or remove invalid response results in the questionnaire.

3. The method according to claim 2, wherein the determining whether the response result is valid according to all response results of the topics corresponding to the response results that do not meet the preset expectation of the prediction result, so as to retain valid response results in questionnaires or reject invalid response results in questionnaires, comprises:

determining the variance of the numbers corresponding to the options of all the reply results according to the numbers corresponding to the options of all the reply results;

and determining whether the response results are valid according to the variance and the tolerance threshold value so as to retain valid response results in the questionnaire or eliminate invalid response results in the questionnaire.

4. The method according to claim 3, wherein said determining whether the response results are valid according to the variance and tolerance threshold to retain valid response results in the questionnaire or to reject invalid response results in the questionnaire comprises:

if the variance is smaller than the tolerance threshold, determining that all preset expected reply results which do not accord with the predicted result are valid, and keeping the valid reply results;

and if the variance is larger than or equal to the tolerance threshold, determining that all preset expected reply results which do not accord with the prediction result are invalid, and removing the invalid reply results.

5. The method of claim 2, further comprising:

if the times are less than or equal to the time threshold, readjusting the order of options in the corresponding questions of the reply results which do not accord with the preset expectation of the prediction result;

and supplementing the topics with the adjusted option order to a to-be-released list in the questionnaire so that the research user answers again.

6. The method according to any one of claims 2 to 5, wherein the answer results of the questions in the questionnaire are of a single-choice type, and the preset expectation is that: hit the first X sets of the prediction result with probability size ordering, wherein X is an integer larger than 0;

if the answer result meets the preset expectation of the prediction result, determining that the answer result of the questionnaire is valid, and keeping the valid answer result comprises:

if the answer result hits the first X sets with the probability size in the predicted result, the answer result is determined to be valid, and the valid answer result is reserved.

7. The method according to any one of claims 2 to 5, wherein the answer results of the questions in the questionnaire are of a multi-choice type, and the preset expectation is that: missing the last Y sets of the probability magnitude ordering in the prediction result, wherein Y is an integer larger than 0;

if the answer result meets the preset expectation of the prediction result, determining that the answer result is valid, and retaining the valid answer result, including:

and if the reply result does not hit the last Y sets of the probability magnitude sequencing in the prediction result, determining that the reply result is valid, and reserving the valid reply result.

8. The method according to any one of claims 2 to 5, wherein the answer results of the questions in the questionnaire are of an open type, and the preset expectation is that: the emotion is consistent with the expression emotion in the prediction result;

if the answer result meets the preset expectation of the prediction result, determining that the answer result of the questionnaire is valid, and keeping the valid answer result, wherein the step of:

segmenting the answer result of the question by using sentence segmentation to obtain a segmentation result;

inputting the target result into a second model to obtain answer emotion, wherein the second model is used for determining the emotion expressed by the investigation user in the open-type questions;

9. The method according to any one of claims 1 to 5, wherein before said obtaining the answer results of the survey user to each topic in the questionnaire, the method further comprises:

determining a user label of a research user based on the user representation;

extracting questions corresponding to the questionnaire labels with consistent user labels from a question bank to generate the questionnaire;

pushing the questionnaire to the survey user.

10. The method according to any one of claims 1-5, wherein before said determining whether the answer result of the user is valid based on the predicted result, the method further comprises:

obtaining game characteristic data of at least one user corresponding to the theme;

and training a machine learning algorithm model according to the game feature data of the at least one user to obtain the first model.

11. A questionnaire processing apparatus, characterized by comprising:

12. An electronic device, comprising: a processor, a memory, and computer program instructions stored on the memory and executable on the processor;

the processor, when executing the computer program instructions, implements a method of processing a questionnaire as defined in any of claims 1 to 10 above.

13. A computer-readable storage medium having stored thereon computer-executable instructions for implementing the method of processing a questionnaire according to any one of claims 1 to 10 when executed by a processor.