TWM576284U - Interactive appraisal system - Google Patents

Abstract

An interactive appraisal system includes a user interface, a chatting robot device and a semantic analyzing engine. The chatting robot device receives a question message from an appraisal database to actively present the question message on the user interface. A reply message is sent to the semantic analyzing engine when the user interface receives the reply message. The semantic analysis engine extracts keywords from the reply message by using an algorithm module so as to parse the reply message. A classification module classifies the analyzed reply message into categories according to a field distribution. A score module gives scores according to contents of the classified reply message. A report generating device then compiles the reply message, the categories and the scores to generate a report.

Description

Interactive check system

This creation is an interactive check system that asks users and chat bots to answer and answer the results for use as check statistics.

The check is used to collect feedback from the user and analyze the feedback for subsequent management. In general, the check is used for course comments on teaching content, follow-up of consumer behavior, feedback on service content or products, and so on.

The existing inspections mainly use paper-based testing or online answering. However, paper-based testing is subject to time and place restrictions in actual application, and the response rate is generally lower. And the paper used in paper has environmental concerns. On the other hand, the paper is based on the lack of interaction with the subjects. Under the same questionnaire, it may be because of different questionnaires. There are no errors in the questionnaire, and there is no way to provide other explanations for the subject of the questionnaire to enable the subject to get different descriptions of the topics through interaction, thus affecting the accuracy of the answer.

Online answering can electronically use the content of the paper questionnaire to have more expressions and ways, and can reduce the time and place of the subject. The use of limitations, but still does not provide the interaction mechanism that is lacking in the paper-based answering process. When the subject cannot understand the problem, the subject cannot be further explained by the interactive mechanism. The online answering cannot be greatly improved. The accuracy of the questionnaire. In addition, in the elaboration of qualitative questions, it is impossible to get more information through the process of interaction.

In order to solve the foregoing problems, the present invention provides an interactive check system, which includes: a user interface; a chat robot device connected to the user interface and the check database, and the problem is obtained by the check database. The message actively presents the problem message to the user interface. After receiving the reply message, the user message transmits the reply message to the semantic analysis engine. The semantic analysis engine includes an algorithm module and a classification module. And the scoring module, wherein the algorithm module extracts the reply message by keyword extraction to parse the reply message, and then transmits the parsed reply message to the classification module and the scoring module for the The classification module classifies the parsed reply message into a category configured according to the domain, and causes the scoring module to give a score according to the content in the classified reply message; and the report generating device, which collects the semantic parsing engine The reply message parsed, the classification result of the classification module, and the score given by the scoring module are used to generate a report.

The interactive check system further includes a reply database that stores the reply message processed by the semantic parsing engine to provide the reply message to the report generating device.

The aforementioned interactive check system further includes a learning program device, The machine learns the result of the semantic analysis engine through the model to transmit the result to the semantic analysis database.

In the foregoing interactive check system, the chat robot device judges the validity of the check according to the time interval in which the user interface continuously receives the reply message.

In the foregoing interactive check system, the chat robot device judges the validity of the check according to the obtained problem message with a reverse intention.

In the foregoing interactive check system, the problem message presented by the chat robot device is presented by a text button, a picture selection or a text answering, wherein the question message of the text button is given by an button. The chat robot device converts the reply message selected by the text button into a text; the question message of the image answering mode provides a selection through the image for the chat robot device to convert the reply message via the picture answer to the text; The question message in the text is to express the question message in text form and receive the reply message in text form.

In the foregoing interactive check system, the algorithm module performs the keyword extraction through the semantic parsing database, and generates the reply message of the multi-dimensional word break to parse the reply message after reducing the dimension.

In the foregoing interactive check system, the classification module classifies the content in the reply message according to the classification model in the semantic analysis database, and the category configured according to the domain is the category data set in the classification module. In the library.

The aforementioned interactive check system, wherein the scoring module is based on the A score is given to the positive or negative range of the emotional vocabulary of the content in the reply message.

The two-way interactive intelligent check system of this creation, under the popularization of the current network and smart mobile devices, the subjects who are checked can be more intuitively interactive without being limited by the time and place of the answer. Dialogue, the verification of interactive chat with chat robots on the instant messaging platform can not only increase the answer rate, but also increase the accuracy of the questionnaire by means of an instant interaction mechanism and a mechanism for judging the validity of the questionnaire during the check process. .

11‧‧‧User interface

12‧‧‧chat robot

13‧‧‧Check database

14‧‧ ‧ semantic analysis engine

15‧‧‧Reply database

16‧‧‧Learning program device

17‧‧‧Report generation device

S1-S14‧‧‧Steps

The first picture is a schematic diagram of the system architecture of the interactive check system of the creation; the second to third figures are the user interface of the creation and the use of the chat robot module; and the fourth picture is the interactive check of the creation Schematic diagram of the operation of the system.

The present invention provides an interactive check system. As shown in FIG. 1 , the interactive check system includes: a user interface 11 , a chat robot device 12 , a check database 13 , a semantic analysis engine 14 , and a reply database 15 . The learning program device 16 and the report generating device 17 are provided.

The user interface 11 is an operation interface for receiving a reply message of a user's voice, text, pictures, options, etc., and can be used by a social software, a communication platform, a website, a chat room, or an application. The user interface 11 communicates or communicates with the chat robot by means of voice input, text input, and clicking.

The chat robot device 12 is connected to the user interface 11 and the verification database 13. The chat robot device 12 obtains the problem message that has been built in the verification database 13 by the verification database 13, and uses buttons, texts, The image, or any combination thereof, actively presents the problem message to the user interface 11. As shown in the second and third figures, after the user interface 11 receives the reply message from the user, the reply message is sent to the semantic analysis. Engine 14. Wherein, when the chat robot device 12 provides a question message of a text button, a picture selection or a text answer, the question message of the text button presents an option in the form of a button or a card, and the user selects an answer according to the option, the chat robot The device 12 converts the reply message of the text button selected by the user into a text and stores it in the reply database 15; the question message of the picture pick-up is provided in different blocks of the picture design to indicate different options, and the user clicks the block to In response to the reply message, the chat bot 12 converts the reply message of the picture block selected by the user into a corresponding text, and stores the corresponding text in the reply database 15; the question message of the text is answered by The chat robot device 12 transmits the question message of the check database 13, and the user answers the reply message in text form, and the chat robot device 12 stores the user's text reply message in the reply database 15; the reply database 15 stores the foregoing The reply message processed by the semantic parsing engine to provide the reply message to the report Set 17.

The semantic analysis engine 14 includes an algorithm module, a classification module, and a scoring module, wherein the algorithm module takes the user's reply message to the natural The language processing algorithm analyzes its semantics, mainly through keyword extraction, through a large amount of text according to the semantics of the lexical part of the semantic analysis database, the text of the reply message is broken, to resolve the reply message, and the same The way to generate vocabulary for model training, as shown in Listing A below, is that the sentence after the word break is a multi-dimensional architecture that makes the statement easier to parse after the reduced dimension.

The classification module of the semantic analysis engine 14 assigns the reply message parsed to the plurality of categories configured according to the industrial field, and the classification module includes a category database and a classification mechanism, wherein the category database is configured with different pre-built configurations. The classification is based on different industrial fields corresponding to the inspection. The classification mechanism is a classification model constructed on the semantic analysis database. It performs preliminary assignment and categorization of related vocabulary of each category, and repeats training classification through a large number of texts. The accuracy of the model is taken as an example of the examination of the learning effectiveness of the education industry. The categories used are "teaching method", "course content" and "examination scoring method". For example, in Table A, the reply message will be Classified into the classification of "teaching methods".

The scoring module of the semantic analysis engine 14 gives a score according to the content in the reply message classified by the classification module, for example, the vocabulary in the content is filled according to the user's text, especially the vocabulary with the emotional meaning defined. Determine the value range, the value range is divided into positive and negative directions, and the magnitude and width of the value domain will be determined according to the inspection requirements of different industrial fields. For example, taking the examination of the learning effectiveness of the education industry as an example, the value range Can be set in the interval of [-1,1], as in the example of Table A above, if the word "uninteresting" is a negative vocabulary in the scoring tag database, the scoring module will be the second picture. "grant The sentence of the class method is classified according to the degree of negative of the emotional vocabulary, for example, "-0.63 points", and the aforementioned numerical values are collected in the report generating device.

Table 1 below is an example of classification and scoring by the above classification module and scoring module:

The following example is an example of the training phase and the prediction phase of the semantic analysis engine 14. In the training phase, the sentence input by the user is subjected to word segmentation and part of speech analysis to obtain the matrix O of the following list 2, and depending on the part of speech, it is possible Deriving a plurality of matrices (for example: O ₁ , O ₂ , ..., O _n ); lowering the dimension of the high-dimensional matrix obtained after the above-mentioned word-breaking to obtain the matrix S of the following list 3 (according to the difference in part of speech) It is possible to derive the matrix S ₁ , S ₂ ,..., S _n ); put the matrix S and the category value C into the classification module to establish a classification model to derive the model A; and then use the value according to each category The matrix S and the score V are input, and the model is built using the scoring module to obtain the model B (depending on the category, it is possible to derive the matrices B ₁ , B ₂ , ..., B _n ).

In the prediction stage, the sentence input by the user is subjected to the word segmentation and part of speech analysis to obtain the aforementioned matrix O (depending on the part of speech, it is possible to derive a plurality of matrices O ₁ , O ₂ , ..., O _n ); The matrix of the high-dimensional matrix obtained after the above-mentioned word-breaking is reduced to obtain a matrix S (depending on the part of speech, it is possible to obtain the matrix S ₁ , S ₂ , ..., S _n ); placing the aforementioned matrix S The model A is acquired to obtain the category value C; the corresponding model B is selected based on the category value C described above, and the predicted score is obtained through the matrix S.

For example: user input: the content of this lesson is really interesting.

Forecast category is: Course content

The forecast score is: 4

The report generating device 17 of the present creation summarizes the reply message of the parsed and broken words, the classification result of the classification module, and the score given by the scoring module from the reply database 15, and after statistics, in the form of text or graph Generate reports and summarize key indicators for decision making through algorithmic modules.

The creation further includes a learning program device 16 that performs machine learning through the model to analyze the results of the semantic analysis engine to transfer the results to the semantic analysis database.

In order to ensure the validity of the checking process and to confirm whether the user actually answers the question, the chat robot device 12 can confirm the time for the user to fill in the reply by the length of time that the user interface continuously receives the reply message. In order to judge the validity of the check, the chat bot 12 can also randomly transmit a problem message with a reverse meaning between the transmission of a plurality of question messages to judge the validity of the check (for example, at In the first question, ask if you like the event, and in the fifth question, ask the user if you don't like the event.

Please refer to the operation flow diagram of the interactive check system in Figure 4. When the user activates the interactive check system of this creation, the interactive check system system judges the user's access to the system S1, and the chat robot device will ask User intention S2, to confirm the identity (if the verification needs to confirm the identity of the user, you need to provide consent binding, if there is no such requirement, do not bind the identity) to determine the user's access to the inspection system. There are three ways to ask for intent to enter the checkout system: provide code to enter the check, go straight to check (for example, with the chat robot, say: I want to answer the OO lecture), and click the button to enter the check (use Chatter device In the interaction, when the chat robot device asks the user's intention, the user's quick reply button is provided for the user to click, and the user is allowed to enter the check S4, and the chat robot device asks for the check category S4. Determining whether there is a temporary storage S5, and checking whether the check has a historical record in the identity according to whether the check has a temporary storage characteristic, and asking whether to continue the last answer S7, and if continuing, continue After the check S8; or after determining that there is no temporary storage content, the chat robot device describes the check content S9, and interacts with the chat robot device to check the process S10. After the interactive check process is finished, the check interaction system confirms whether to submit. Replying to the message S11, if yes, importing the database S12, if not, the chat robot device inquires about the intention S2; then after S12, asks whether to continue to answer the other question S13, and if so, returns to step S4, if If no, the current check S14 is ended; if it is asked whether to continue the last reply S7 or whether to submit the reply message S11, the answer back to the chat robot device will ask the user Intent S2.

The detailed description above is intended to be illustrative of a possible embodiment of the present invention, and is not intended to limit the scope of the present invention. Any equivalent implementations or modifications that are not departing from the spirit of the present invention should be included in The patent scope of this case.

To sum up, this case is not only innovative in terms of technical thinking, but also has many of the above-mentioned functions that are not in the traditional equipment used by the law. It has fully complied with the statutory creation patent requirements of novelty and progressiveness, and applied for it according to law. Approved this patent application for creation, in order to encourage creation, to the sense of virtue.

Claims (9)

An interactive check system includes: a user interface; a chat bot device connected to the user interface and configured to obtain a problem message to actively present the question message to the user interface for use After receiving the reply message, the interface transmits the reply message; the semantic analysis engine includes an algorithm module, a classification module and a scoring module, wherein after receiving the reply message, the algorithm module will reply The message is extracted by the keyword to parse the reply message, and the parsed reply message is transmitted to the classification module and the scoring module, so that the classification module classifies the parsed reply message to be configured according to the domain. a category, and causing the scoring module to give a score based on the content in the reply message after the classification; and a report generating device that collects the reply message parsed by the semantic parsing engine, the classification result of the classification module, and the Score the score given by the module to generate a report. The interactive checking system of claim 1 further includes a replying database for storing the reply message processed by the semantic analysis engine to provide the reply message to the report generating device. According to the interactive inspection system of claim 1, further comprising a learning program device that performs machine learning through the model to analyze the result of the semantic analysis engine to transmit the result to the semantic analysis Library. According to the interactive inspection system of claim 1, wherein the chat robot device judges the validity of the check according to the time interval in which the user interface continuously receives the reply message. According to the interactive inspection system of claim 1, wherein the chat robot device judges the validity of the check according to the obtained problem message with a reverse intention. According to the interactive inspection system of claim 1, wherein the problem message presented by the chat robot device is presented by a text button, a picture selection or a text, wherein the text button is The question message is an option given by the button for the chat robot device to convert the reply message selected by the text button into a text; the question message of the picture answer is provided through the picture for the chat robot device to select the picture via the picture The reply message is converted into text; and the question message that the text answers is to express the question message in text form and receive the reply message in text form. According to the interactive checking system of claim 1, wherein the algorithm module performs the keyword extraction by using the semantic analysis database to generate the reply message of the multi-dimensional word segmentation, so as to After the dimension is lowered, the reply message is parsed. According to the interactive inspection system described in claim 1, wherein the classification module is based on a classification model in a semantic analysis database. The content in the reply message is classified, and the category configured according to the domain is set in the category database of the classification module. The interactive checking system according to claim 1, wherein the scoring module gives a score according to a positive or negative value range of the emotional vocabulary of the content in the reply message.