KR102211537B1 - Interview supporting system - Google Patents

Abstract

The present invention relates to an interview support system, and more particularly, by matching a keyword selected from the user's history information with a question item in a question DB to generate a user question item, it is possible to perform customized interview coaching, and the user history If the information is written to the extent that the number of user keywords cannot be secured, the user history information and the history information of a third party with a high keyword matching rate are provided as a reference sample, or keyword reinforcement is possible from the history information of a third party, It is possible to reduce the amount of time the user spends on supplementing the history information, and by allowing the user question items with high meaning similarity to the actual interview items to be exposed first, the user can prioritize the question items with high probability of the questions. , If the keyword selected from the user's history information satisfies the number of the set range, it is provided to the user without sorting the user question items, allowing prompt provision of the question item, and the current topic issue can be applied to the user question item. Therefore, it relates to an interview support system that can quickly respond to an interview with a fluid issue.

Description

Interview supporting system

The present invention relates to an interview support system, and more particularly, by matching a keyword selected from the user's history information with a question item in a question DB to generate a user question item, it is possible to perform customized interview coaching, and the user history If the information is written to the extent that the number of user keywords cannot be secured, the user history information and the history information of a third party with a high keyword matching rate are provided as a reference sample, or keyword reinforcement is possible from the history information of a third party, It is possible to reduce the amount of time the user spends on supplementing the history information, and by allowing the user question items with high meaning similarity to the actual interview items to be exposed first, the user can prioritize the question items with high probability of the questions. , If the keyword selected from the user's history information satisfies the number of the set range, it is provided to the user without sorting the user question items, allowing prompt provision of the question item, and the current topic issue can be applied to the user question item. Therefore, it relates to an interview support system that can quickly respond to an interview with a fluid issue.

Thanks to the influence of the Internet, online education programs using the Internet are gaining popularity and widespread use in the education world. In most online education systems, lectures by lecturers in the subject are produced as video files and uploaded on the Internet through an education providing server, and then users download and watch the lectures they want regardless of time and place using a PC connected to the Internet. It provides on-demand lecture training materials in a streaming method that you can learn.

These existing online education systems allow users to download and play educational materials from education-related websites such as cyber universities and Internet education companies, and learn at home on the multimedia Internet. Downloaded educational materials are available anytime, anywhere through a PC. Repetitive learning is possible, and it includes a function that allows users to send and receive questions about educational materials and lecture evaluations with educational service providers through the Internet.

However, although such an existing online education system has the advantage of being able to receive education at a time and place he wants, it has a limitation that most of the lecture contents are limited to the one-sided delivery of knowledge of the lecturer about the subject.

In addition, the conventional online education system focuses only on the transfer of knowledge of subjects such as Korean, English, and mathematics. However, in recent years, the importance of interviews for college entrance examinations and company employment is increasing, and since interview questions are asked from the contents of personal history and self-introduction letter, personalized interview guidance is not possible only by transferring the existing knowledge. There were no problems.

Therefore, it is necessary to develop a customized interview support system suitable for personal history information.

KR10-1671179 (registration number) 2016.10.26.

An object of the present invention is to provide an interview support system capable of personalized interview coaching by matching a keyword selected from user history information with a question item in a question DB to generate a user question item.

In addition, the present invention provides the user history information and the history information of a third party having a high keyword matching ratio as a reference sample when the user history information is written to the extent that the number of user keywords cannot be secured, or the history of a third party It is an object of the present invention to provide an interview support system capable of reducing the amount of time a user spends on supplementing history information by reinforcing keywords from information.

In addition, an object of the present invention is to provide an interview support system that enables users to prioritize question items with a high probability of questioning by first exposing user question items having a high meaning similarity to the previously written interview items. .

In addition, an object of the present invention is to provide an interview support system capable of promptly providing question items by providing them to a user without alignment of user question items when the number of keywords selected from the user's history information satisfies the set range. .

In addition, an object of the present invention is to provide an interview support system capable of promptly responding to a fluid issue by applying an issue that is currently hot topic to a user question item.

The present invention provides a question DB in which a question item or a common question item from which keywords are excluded is stored; A history input unit for inputting user history information from a user; A keyword selection unit for selecting at least two user keywords from the user history information; A question matching unit extracting each question item matching each of the user keywords from the question DB; A question generator for completing a user question item by combining the question item extracted from the question matching unit and the user keyword; And a question transmission unit for providing the user question item or the common question item to a user.

In addition, the present invention, a history DB for storing by matching the user keyword with the user history information; And a third-party history information selection unit for selecting third-party history information in which at least two or more identical keywords match the user keyword from among the third-party history information stored in the history DB when the number of user keywords is less than the set number; .

In addition, the present invention includes a history information providing unit for providing the third party history information selected from the history information selection unit to the user.

In addition, the present invention includes a keyword reinforcement unit for providing the stored keyword matched to the third party history information selected from the history information selection unit to the user or the question matching unit.

In addition, the present invention includes a history information classification unit that divides the user history information into a plurality of history sections, wherein the history DB matches and stores the user keyword with the history section information, and the third party The history information selection unit selects third-party history information so that the third-party history information having more keywords matched with the same history section information as the user keyword becomes priority, or, in the history section information identical to the user keyword. The third-party history information is selected so that the third-party history information having a higher sum total by multiplying the matched keywords by a preset weight for each history section becomes a priority.

In addition, the present invention, the actual interview DB in which the previously recorded interview items actually recorded; A question sorting unit that compares the user question items generated by the question generation unit with the existing interview items and arranges the user question items having a higher semantic similarity and a higher overlap rate to be priorities, wherein the question transmission unit includes the User question items are transmitted to the user in a sorted order from the question sorting unit.

Further, the question sorting unit of the present invention skips the sorting process of the user question items when the number of the user keywords is within a set range.

In addition, the present invention, an issue collection unit for collecting and storing the issue keyword in real time; Issue keywords that are duplicated in the issue collection unit are ranked according to the number of duplicates collected in units of a set time, but if the period maintained above the set ranking lasts for more than a set time or the number of exposures above the set ranking is more than the set number, key issue keywords An issue selection unit that selects as; and the question matching unit extracts a question item matching the main issue keyword from the question DB.

In the present invention, by matching a keyword selected from the user's history information with a question item in a question DB to generate a user question item, it is possible to perform customized interview coaching.

In addition, the present invention provides the user history information and the history information of a third party having a high keyword matching ratio as a reference sample when the user history information is written to the extent that the number of user keywords cannot be secured, or the history of a third party By reinforcing keywords from information, there is an effect of reducing the time the user spends on supplementing and creating history information.

In addition, the present invention has an effect of allowing a user to first prepare a question item having a high probability of submission by first exposing a user question item having a high meaning similarity to an existing interview item actually written out.

In addition, according to the present invention, when a keyword selected from the user's history information satisfies the number of a set range, it is possible to provide a quick question item by providing it to the user without sorting the user question items.

In addition, since the present invention can apply the currently topical issue to the user question item, it is possible to quickly respond to an interview with a fluid issue.

1 is a conceptual diagram of an interview support system according to an embodiment of the present invention.
2 is a flowchart of an interview support system according to an embodiment of the present invention.
3 is a flowchart of a question sorting unit of an interview support system according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The present invention, as shown in Figs. 1 to 3, a question DB 10 in which a question item or a common question item from which keywords are excluded is stored, a history input unit 21 for inputting user history information from a user, and , A keyword selection unit 22 for selecting at least two user keywords from user history information, a question matching unit 11 for extracting each question item matching each user keyword from the question DB 10, and a question A question generation unit 12 that combines the question item extracted from the matching unit 11 and a user keyword to complete a user question item, and a question transmission unit 13 that provides a user question item or a common question item to the user. It is composed by

Prior to the description of the present invention, each unit to be described below may be utilized as a function embodied in a separate device, or may be software or hardware implemented in a device such as one server device, and for the description of the present invention For convenience, only functional names are used, and it is clarified that each part does not mean one independent entity as the name. For example, in the case of a word processor system, the input unit may be a keyboard or a mouse, but it may also mean a file loading function inside the word processing software, and it states that'the word processor system receives data through the input unit.' It is obvious to those skilled in the art that the keyword is not input only by typing on the keyboard. That is, a series of processes in which certain data is consequently provided to the next processing step is referred to as'input' in the present invention.

The question DB 10 serves to store question items from which keywords are excluded. These questions DB (10), for example, "What do you think is the greatest virtue that () should have?", "Please tell me why the () activities in high school should be reflected in the college admissions process." A number of question items excluding keywords are stored in the form of. In question items from which such keywords are excluded, the user keywords selected by the keyword selection unit 22 are filled in the blanks. In this case, the question matching unit 11 to be described below determines which user keyword is applied to the question item from which which keyword is excluded.

In addition, the question DB 10 also stores common question items. Common question items are question items that do not need to be matched with user keywords. For example, "What do you think is the problem with early English education?", "What should I do to develop logical thinking?" It has the same shape as

On the other hand, the question DB 10 may also be linked with the actual interview DB 30 to be described below. A common question item is extracted and stored from the previously interviewed items stored in the actual interview DB 30, or, or the actual interview DB 30 ) Can be extracted and stored in a form in which keywords are excluded from the previous interview items.

The history input unit 21 serves to receive user history information from a user, and may be in the form of a web bulletin board completed by receiving data through typing of a keyboard, or a window to upload a completed word file or the like.

The history input unit 21 is in a state in which the user history information exists in a form recognizable by the system of the present invention, that is, in the form of text in a standardized form, so that the history information classification unit 23 or the keyword selection unit 22 can process it. You just need to be satisfied. In addition, the user history information input through the history input unit 21 is stored in the history DB 20 so that other users can use it as third party history information.

When the input user history information does not satisfy the number of characters required in the form, the history input unit 21 may request supplementation thereof.

The keyword selection unit 22 serves to select at least two user keywords from the user history information input from the history input unit 21. This is done by analyzing the morpheme of the sentence, and tagging the part of speech of each word included in the sentence in the user history information. Part of speech tagging refers to the process of allocating correct part of speech information to each word according to the context in which the word is used within a sentence.

Part of speech tagging methods include a rule-based part-of-speech tagging method and a statistics-based part-of-speech tagging method commonly used in natural language processing. The statistical-based approach is the probability or uncertainty obtained by analyzing a large amount of raw or tagged corpus including real-world natural language examples and accompanying information, and extracting statistical information about natural language. It refers to a method of probabilistically solving a lexical ambiguity problem using (uncertainty). In addition, the rule-based approach refers to a method of finding a common principle or decisive rule applied to POS tagging, and decisively solving lexical ambiguity by using this. The system of the present invention can tag parts of speech including both a rule-based approach method, a statistics-based approach method, and a public part-of-speech tagging method.

Part of speech corresponding to a noun among the tagged parts of speech can be classified as a user keyword and a basic meaning corresponding thereto may be tagged. At this time, the meaning of the word may be classified by referring to a predetermined Word Sense database in which the general meaning of the word is recorded.

For example, "I was active in a soccer club in high school, and through this, I was able to learn how to properly coordinate individual competencies with the team's organizational abilities." The sentences "high school", "time", "football", "Club", "In", "Activity", "Did", "This", "Through", "Individual", "Of", "Competency", "Eul", "Team", "Of", "Organizational Power It can be ", "and", "appropriately", "tuning", "doing", "how to", "to", "learning", "do", "can", "was". Among these nouns are "high school", "soccer", "club", "activity", "individual", "capability", "team", "organizational skills", "coordination", "method" as nouns and stored in user keywords Can be. At this time, if you reselect only meaningful keywords, "high school", "soccer", "club", "individual", "capability", "team", "organizational skills", "coordination" can be the targets. , Going further from here, "soccer" and "club" can be grouped together as related keywords, "individual" and "team", "individual" and "organizational power", "team" and "organizational power", "team" and " Coordination", or "organizational power" and "coordination" may be stored as keywords associated with each other. Such a related keyword may be used when comparing a user keyword and a keyword matched with the third party's history information in the third party history information selection unit 24 in the future.

Meanwhile, user history information can be largely composed of two forms. One is in the form of a school life detail record, and the other is in the form of a self-introduction letter. In the case of the school life record book, there are regularized forms such as personal information, academic records, attendance status, awards, qualifications, career hopes, creative experience activity status, subject learning development status, reading activity status, etc. In the case of self-introduction, it is again divided into college admissions self-introduction and employment self-introduction. The college self-introduction has a relatively free form, such as the effort spent on learning, learning experiences, and learning and feelings. It is largely divided into two forms, including the experience achieved, the aspirations after entering the company, or the background of growth, the motivation for applying, the strengths and weaknesses of the personality, and the aspirations after entering the company.

Therefore, the user history information for college admission is the school life record book and self-introduction letter, and the user history information for employment is a resume and self-introduction. In the case of the self-introduction letter, the forms are different between college and employment, but they differ greatly in their uses. It is showing a form that is not.

Using this, user history information can be divided into a plurality of history sections.In the case of a self-introduction letter for college admission, three sections of the effort devoted to learning, learning experience, and learning and feelings, growth background in the case of employment self-introduction letter, and strength of work Motivation), experience of achieving a sense of accomplishment (strengths and weaknesses of character), and aspirations after joining the company. In the case of school life records or resumes, it can be divided into several categories according to the division point of the written form.

The history information classifying unit 23 divides the user history information into a plurality of history sections. The user history information divided into a plurality of sections is selected by the keyword selection unit 22 for each section, and the history DB 20 matches and stores the user keywords with the history section information.

The question matching unit 11 serves to extract from the question DB 10 each question item that matches each user keyword selected by the keyword selection unit 22. Among the question items, in the question items from which the keyword is excluded, information on which category of the user keyword should be inserted in the blank, that is, the part from which the keyword is excluded, is designated. For example, "How did your () activities help you grow physically or mentally?" In the blank of the question item, a user keyword grouped with a club activity and a related keyword may be matched. Alternatively, a user keyword selected from the reading activity situation of the school life record may be matched to the blank of the question "What is the criterion for selecting the book when reading ()?"

The question generation unit 12 completes the user question item by combining the user keyword and the question item from which the keyword extracted from the question matching unit 11 is excluded. This is a function that can be performed directly by the question matching unit 11, but the question generating unit 12 goes further from here to insert the user keyword into the blank of the question item, as well as select an appropriate part of speech suitable for the user keyword to be inserted. To provide. To this end, the question generator 12 may correct the spelling of each phrase through the spell check module, and thus, not only a single user keyword, but also a plurality of user keywords grouped with related keywords may be generated and provided as sentences.

For example, "Soccer" was selected as a user keyword matching the question item "How did () activities help high school life?", and "Club", a user keyword related to this, was also selected. It can be generated as a sentence of "How did the activity help you in high school?"

The question transmission unit 13 serves to provide a user question item or a common question item generated by the question generation unit 12 to the user. The question transmission unit 13 may be a general network communication unit, and if the system of the present invention is configured as an integrated terminal, a monitor, printer, or other output device may correspond to this.

On the other hand, there may be a case in which the number of keywords selected by the keyword selection unit 22 by using user history information does not reach the set number. The number of settings is the minimum number of user keywords for creating user question items. For example, 50 user question items must be created, but if the total number of selected user keywords is 30, 20 shortfalls occur. do. The main reason for this is that the user history information does not contain enough information, and if the self-introduction letter, which requires 1,500 characters, is filled in only 500 characters, the selection of user keywords is very likely to be less than half.

As described above, when the number of user keywords is less than the set number, the system of the present invention has to allow the user to supplement the user history information or add another keyword of a similar series in addition to the keyword selected from the user history information to complete the user question item.

To this end, a third party history information selection unit 24 is configured to select third party history information in which at least two or more identical keywords are matched with a user keyword among the history information of a third party stored in the history DB 20.

In this case, since third-party history information including all user keywords is rarely present, third-party history information having a high similarity should be selected as the next best option.

Therefore, in the present invention, the history section information described above is utilized. The user keyword is matched with the history section information and stored in the history DB 20, which is also stored by matching the third-party history section information with the third-party keyword. That is, even if the user keyword and the third-party keyword are the same word, if they belong to different history sections, they are treated as different words. Through this, the third-party history information selection unit 24 of the present invention selects the third-party history information in the order in which keywords matched with the same history section information as the user keyword are found.

On the other hand, the history sections may not have the same weight. For example, since information such as personality strengths and weaknesses and motivation for applying is more likely to be used as more important information during an interview than information such as academic background information and growth environment, questions related to the questions provided to users may also be more efficient. have. Therefore, the third-party history information selection unit 24 multiplies the keywords in the third-party history information matched with the same history section information as the user keyword by a preset weight for each history section, and the sum of the third-party history is in the highest order. Select information. Accordingly, although there are fewer keywords than the third-party history information overlapping the user keyword, the third-party history information, which is a history section having a high weight, in the history section overlapping the user keyword can be selected.

Third party history information selected in this way can be used in two ways.

First, it is provided to the user and may request supplementation of user history information. To this end, in the present invention, the history information providing unit 25 is configured, and the history information providing unit 25 provides the user with the third party history information selected from the history information selection unit. This is the meaning of providing a sample, and it is preferable to limit the provided third party history information to the self-introduction letter with the consent of the user who entered the corresponding history information. The user can see the third-party history information having a keyword similar to that of the user, and realize the part that he or she has written insufficiently, and can re-enter the supplemented user history information. This means that most users are having difficulty in writing self-introduction letters, and even if you refer to someone else's self-introduction letter, there are many things in your field of interest or in a completely different field than the relevant field. This is to prevent.

Second, the keyword may be directly provided to the question matching unit 11 and used to generate user question items. To this end, in the present invention, a keyword reinforcement unit 26 is configured to provide a keyword matched to third party history information selected from the history information selection unit and stored to the user or question matching unit 11.

The keyword reinforcement unit 26 provides the user with a keyword of third party history information selected from the history information selection unit. The user may view the keywords expressed in the form of a list and select a keyword corresponding to the user among them, and the keyword reinforcement unit 26 provides the keyword selected by the user to the question matching unit 11 as a user keyword.

Alternatively, the keyword reinforcement unit 26 directly provides the keyword of the third-party history information selected from the history information selection unit to the question matching unit 11. This has a quick advantage in that the user does not go through a separate selection process, but the wrong keyword may be involved in the creation of the user question item. Accordingly, the keyword reinforcement unit 26 compares the keyword of the third party history information selected from the history information selection unit with the user keyword selected from the user history information, and provides it directly to the question matching unit 11 only when the similarity is greater than or equal to the set ratio. Can be configured to For example, the user keywords selected from user history information are "Park Ji-sung", "Soccer", "European Football League", "World Cup", and keywords matched with third-party history information are "Son Heung-min", "Soccer", In the case of "European Football League", "Asian Games", "Dream", "Passion", and "Challenge", it is judged that the keyword matched to the third party's history information can be directly provided to the question matching unit 11 will be.

On the other hand, the actual interview may change according to a specific trend as the times or years change. Therefore, it will be an interview support system that can be used in an actual interview only when the existing questions in the actual interview are applied to the system.

Accordingly, in the present invention, the actual interview DB 30 is configured in which the actual interview items that have been actually previously recorded are stored. Further, a question sorting unit 31 is further configured to compare the user question items generated by the question generation unit 12 with the existing interview items and arrange them in an order having a high semantic similarity and a high overlap rate.

The question arranging unit 31 extracts keywords from the existing interview items and compares them with keywords used in the user question items, and gives a higher order as the number of keywords having similar meanings of words increases by referring to the word sense database. At this time, the keywords extracted from the previous interview items are extracted from all the previous interview items, and the number of duplicates is matched, and how many times the keywords used to determine the similarity among the user question items of the same ranking sorted by the meaning similarity of the words were duplicated. By re-evaluating, user question items including keywords with a large number of duplicates are sorted with higher priority.

The user question items arranged in order from the question sorting unit 31 are transmitted from the question transmission unit 13 to the user according to the corresponding order.

The sorting of user question items by the question sorting unit 31 can be effectively utilized when the number of user keywords selected from user history information is greater than the set number. For example, when the number of appropriate user keywords is 50, and 80 user keywords are selected from user history information, user question items may be generated more than necessary by too many user keywords. In this case, if the user question items are not sorted, the user spends time on the user question items of low importance, and in some cases, the important user question items may have to be moved to the interview without being confirmed.

Therefore, by arranging the user question items from the question arranging unit 31 in a similar order to the existing interview items, that is, placing the user question items that are likely to be actually submitted on the leaflet, the user can more efficiently check the user question items. There will be.

In addition, in this regard, when the user keywords selected from the user history information are within the set error range from the set number, for example, if the set number is 50 and the error range is 5, the number of selected user keywords is 47 The user question item may be provided to the user as it is without arranging the user question items from the question arranging unit 31. Therefore, it is possible to provide a user question item to the user more quickly by excluding unnecessary calculation processes in the number of user keywords in the set range.

In addition, some of the existing interview items may be used as question items from which keywords are excluded, and preferably, other keywords with the highest number of duplicates among the keywords extracted from the previous interview items with high keyword similarity are maintained, and the number of duplicates is reduced. After processing the fewest keywords in the blank, they are registered in the question DB 10 as question items from which keywords are excluded. For example, "How did music club activities affect you?", "How did football club activities affect you?", "How did reading club activities affect you?" When there is an existing interview item called, "Club", "Activity", "Self", and "Influence" are the keywords with the highest number of duplicates, and "Music", "Soccer", and "Reading" are the keywords with the least number of duplicates. Becomes. Therefore, when these previously mentioned interview items are registered in the question DB 10, the keyword “() club activity has had an effect on them?” is registered as a question item excluding the keyword.

On the other hand, some of the previous interview items may be used as common question items. This is when all keywords extracted from the previous interview items with high keyword similarity belong to the keywords of each history information stored in the history DB (20). It is possible. For example, "How did your music club activities affect you?" The previously interviewed item of is not matched by the word “football” with the history information of the user who has been active in the soccer club. However, "How do you plan to develop your talent in connection with your major after entering college?" is more likely to belong to the keywords of each history information more commonly.

However, the present invention is not limited thereto, and it is of course possible for the administrator to directly refer to an existing interview item to create a question item and a common question item from which keywords are excluded and register it in the question DB 10.

On the other hand, the question at the interview is not limited to the user's history information. The question at the interview is actually a method of asking about various issues such as society, culture, politics, and so on, and for this purpose, in the present invention, an issue collection unit 40 and an issue selection unit for generating question items for currently emerging issues (41) is composed.

The issue collection unit 40 serves to collect and store issue keywords in real time, and to this end, analyze text uploaded in real time from news and communities on the Internet and store keywords. At this time, collecting real-time keywords from domestic portal sites may lead to disputes such as traffic issues and copyright issues, so it is possible to conclude an agreement with each portal site for keyword collection or use media such as Google.

The issue keywords collected and stored by the issue collection unit 40 are selected as major issue keywords by the issue selection unit 41. The issue selection unit 41 ranks the issue keywords that are repeatedly collected in the issue collection unit 40 in units of a set time according to the number of overlaps, but the period maintained above the set ranking lasts for a set time or more, or When the number of exposures is greater than or equal to the set number, the relevant issue keyword is selected as the main issue keyword.

For example, if the keyword "Son Heung-min" is in the top 100 ranking according to the number of duplicates per month, and the keyword "Gun-immunity" is listed 10 or more times in the top 100 ranking according to the number of duplicates per month, "Son Heung-min" And "gun exemption" are selected as key issue keywords. Such a set priority, a maintained period, and the number of exposures can be appropriately selected by the administrator.

On the other hand, in addition to the one-off issues such as "Son Heung-min's military exemption", which has been a hot topic since the Asian Games, there may be issues that are talked about over a year or years. For example, "National Affairs Appreciation", which is repeated every year, or "President Trump," which is a hot topic. These mid- to long-term issues are more likely to be questioned in actual interviews compared to one-off issues. Therefore, the issue selection unit 41 puts more weight on the issue keywords that are steadily repeated throughout the month or year, and the number of times the number of exposures above the set ranking to the issue collection unit 40 is more than the set number of times, and the exposure frequency is It is desirable to be configured to be selected as.

The main issue keywords selected by the issue selection unit 41 are provided to the question matching unit 11, and the question matching unit 11 extracts the question items matching the main issue keywords from the question DB 10 to ask questions. It is provided to the generation unit 12.

The present invention having the above-described configuration generates a user question item by matching a keyword selected from the user's history information with a question item in the question DB 10, thereby enabling customized interview coaching.

In addition, the present invention provides the user history information and the history information of a third party having a high keyword matching ratio as a reference sample when the user history information is written to the extent that the number of user keywords cannot be secured, or the history of a third party By reinforcing keywords from information, there is an effect of reducing the time the user spends on supplementing and creating history information.

In addition, the present invention has an effect of allowing a user to first prepare a question item having a high probability of submission by first exposing a user question item having a high meaning similarity to an existing interview item actually written out.

In addition, according to the present invention, when a keyword selected from the user's history information satisfies the number of a set range, it is possible to provide a quick question item by providing it to the user without sorting the user question items.

In addition, since the present invention can apply the currently topical issue to the user question item, it is possible to quickly respond to an interview with a fluid issue.

10: Question DB 11: Question matching unit
12: question generation unit 13: question transmission unit
20: history DB 21: history input unit
22: keyword selection unit 23: history information division unit
24: third party history information selection unit 25: history information providing unit
26: keyword reinforcement part 30: actual interview DB
31: question sorting unit 40: issue collecting unit
41: Issue selection department

Claims (8)

Question DB in which question items or common question items from which keywords are excluded are stored;
A history input unit for inputting user history information from a user;
A keyword selection unit for selecting at least two user keywords from the user history information;
A question matching unit extracting each question item matching each of the user keywords from the question DB;
A question generator for completing a user question item by combining the question item extracted from the question matching unit and the user keyword;
A question transmission unit for providing the user question item or the common question item to a user;
A history DB matching and storing the user keyword with the user history information;
A third-party history information selection unit for selecting third-party history information in which at least two or more identical keywords match the user keyword from among the history information of a third party stored in the history DB when the user keyword is less than a set number;
A keyword reinforcement unit configured to provide a keyword matched to the third party history information selected by the history information selection unit and stored to the user or the question matching unit;
A history information classifying unit that divides the user history information into a plurality of history sections;
Including,
The keyword selection unit selects a user keyword for each of the divided history sections,
The history DB matches and stores the user keywords selected for each history section by matching history section information, and stores the third-party keyword by matching with the third-party history section information,
The third-party history information selection unit may select third-party history information such that the third-party history information having more keywords matched with the same history section information as the user keyword becomes priority, or, the same as the user keyword. An interview support system that selects third-party history information so that the sum of the third-party history information with a higher sum of the keywords in the third-party history information matched with the history section information is multiplied by a preset weight for each history section.
delete delete delete delete The method of claim 1,
Actual interview DB that stores the previously submitted interview items;
A question sorting unit that compares the user question items generated by the question generation unit with the existing interview items and arranges the user question items having a higher semantic similarity and a higher redundancy rate as a priority;
Including,
The question transmission unit interview support system for transmitting the user question items to the user in a sorted order from the question arrangement unit.
The method of claim 6,
The question sorting unit is an interview support system for omitting the sorting process of the user question items when the number of the user keywords is within a set range.
The method of claim 1,
An issue collection unit that collects and stores issue keywords in real time;
Issue keywords that are duplicated in the issue collection unit are ranked according to the number of duplicates collected in units of a set time, but if the period maintained above the set ranking lasts longer than the set time or the number of exposures above the set ranking exceeds the set number The issue selection department selected as;
Including,
The question matching unit, an interview support system for extracting a question item matching the main issue keyword from the question DB.

Images