JP7818300B2 - Selection equipment and programs - Google Patents

Description

The present invention relates to a selection device and program.

Online interviews using networks have traditionally been conducted. Online interviews can be conducted without the need to physically visit an interview venue, saving both applicants and businesses time and the hassle of reserving a venue. One device proposed for such online interviews is one that can send and receive questions and answers to a terminal carried by the applicant (see, for example, Patent Document 1).

Japanese Patent Application Laid-Open No. 2003-263537

When interviewing a large number of candidates, such as for new graduate recruitment, it was necessary to narrow down the applicants from thousands or tens of thousands of applicants. This required a lot of time and effort to select the successful candidates. Businesses also wanted to hire applicants who best matched their desired talent profile. For example, they wanted to hire applicants who fit the company culture, not just those with good test scores. Furthermore, when using video provided by applicants online, there was the problem of time-consuming selection processes, as all of the video had to be viewed.

The present invention was made in consideration of the above-mentioned conventional situation, and aims to provide a selection device and program that makes it easier to determine successful applicants and can identify applicants who better meet the employer's preferences.

The present invention relates to a selection device that selects applicants for a specified business based on video images containing the applicants as subjects, and includes a video image acquisition unit that acquires the video images, a recommendation level determination unit that determines the pass recommendation level for the applicant based on the features of the applicant contained in the video images, video images containing other applicants as subjects, and information regarding the pass/fail decisions of the other applicants, and an output unit that outputs the pass recommendation level for the applicant based on the determination results.

It is also preferable that the selection device further includes an extraction unit that extracts features of the applicant contained in the video.

Furthermore, it is preferable that the extraction unit labels the applicant based on the extracted features, and the recommendation level determination unit determines the pass recommendation level of the applicant based on the label assigned to the applicant, the labels assigned to the other applicants, and information regarding the pass/fail decisions of the other applicants.

The selection device also preferably includes an evaluation unit that evaluates the tendency of the judgment results regarding the pass recommendation level, and an adjustment unit that adjusts the feature amounts that are emphasized based on the evaluation results, and the recommendation level judgment unit determines the pass recommendation level using the adjusted feature amounts.

It is also preferable that the extraction unit extracts at least one feature of the applicant's clothing, facial expression, tone of voice, ambient lighting, room background, facial features, body type, hairstyle, and shooting environment contained in the video.

Furthermore, it is preferable that the recommendation level determination unit determines the applicant's recommendation level for passing based on labels attached to video images of past successful applicants.

Furthermore, it is preferable that the recommendation level determination unit determines the recommendation level for passing the applicant based on feature quantities related to at least one of the logic of the successful applicant's responses, speaking style, behavior, statements, and the sincerity of the statements.

Furthermore, it is preferable that the recommendation level determination unit further determines the pass recommendation level based on information regarding the applicant's status and external evaluation factors.

It is also preferable that the selection device further includes a judgment unit that determines whether an applicant who has been judged to have a predetermined or higher degree of recommendation for passing has exceeded the passing grade.

It is also preferable that the selection device further includes a fulfillment calculation unit that calculates the applicant's fulfillment level with respect to the label.

It is also preferable that the selection device further includes a simulation execution unit that uses the acquired video image to simulate a pass/fail decision for the specified business, and that the output unit outputs the simulation results.

It is also preferable that the selection device further includes an analysis unit that analyzes the tendencies of the determined successful applicants.

It is also preferable that the selection device further includes an analysis unit that analyzes the tendencies of applicants who are determined not to have passed the pass mark.

It is also preferable that the extraction unit sorts the acquired multiple videos using predetermined features.

The present invention also relates to a selection device that selects applicants for a specified business based on video images containing the applicants as subjects, and that includes a video image acquisition unit that acquires the video images, an extraction unit that extracts feature amounts of the applicants contained in the video images, and an output unit that outputs the extracted feature amounts.

The present invention also relates to a program that causes a computer to function as a selection device that selects applicants for a specified business based on video images containing the applicants as subjects, and that causes the computer to function as a video image acquisition unit that acquires the video images, a recommendation level determination unit that determines the degree of recommendation for passing the applicant based on the features of the applicant contained in the video images, video images containing other applicants as subjects, and information regarding the pass/fail judgments of the other applicants, and an output unit that outputs the degree of recommendation for passing the applicant based on the judgment results.

This disclosure makes it possible to provide a selection device and program that makes it easier to determine successful applicants and can identify applicants who better meet the operator's preferences.

1 is a schematic diagram showing the system configuration of a selection system including a selection device according to an embodiment of the present invention. 1 is a block diagram showing a configuration of a selection device according to an embodiment; 10 is a screen diagram showing the results determined by a recommendation level determination unit of the selection device of the embodiment. FIG. 10 is an example of a screen diagram showing the results of evaluation by the evaluation unit of the selection device of the embodiment; 10 is another example of a screen diagram showing the results of evaluation by the evaluation unit of the selection device of the embodiment. 10 is a flowchart illustrating the operation of the selection device according to an embodiment.

A selection device 1 and a program according to an embodiment of the present invention will be described below with reference to FIGS. 1 to 6. FIG.
First, a selection system 100 including a selection device 1 according to this embodiment will be described.

The selection system 100 is a system that determines, for example, the recommendation level for applicant U who has applied for a job at a business (company). The selection system 100 determines the recommendation level for applicant U based on, for example, video images provided by applicant U. In particular, the selection system 100 assigns a high recommendation level to applicant U who is thought to be a good fit for the business's corporate culture. This selection system 100 is a system that is used, for example, in the document screening stage before the first interview. The selection system 100 includes a user terminal 20 and a selection device 1.

The user terminal 20 is, for example, a mobile terminal such as a smartphone or a personal computer. The user terminal 20 is configured to be able to capture moving images of the applicant U. For example, the user terminal 20 captures moving images of the user's responses to tasks (such as speeches) provided in advance.

The selection device 1 is, for example, a server. The selection device 1 is connected to a user terminal 20, for example, via an internet connection N. The selection device 1 is configured to be able to acquire video images captured by the user terminal 20. In other words, the selection device 1 is configured to be able to acquire video images of applicants U applying to a specified business. The selection device 1 selects applicants U for the specified business based on video images that include applicants U as subjects.

Next, the configuration of the selection device 1 according to this embodiment will be described with reference to FIG.
The selection device 1 includes a video acquisition unit 101, a video storage unit 102, an extraction unit 103, a label storage unit 104, a feature storage unit 105, a recommendation level judgment unit 106, a judgment result storage unit 107, an evaluation unit 108, an adjustment unit 109, a judgment unit 110, a satisfaction level calculation unit 111, an analysis unit 112, and an output unit 113.

The video image acquisition unit 101 is realized, for example, by the operation of a CPU. The video image acquisition unit 101 acquires video images transmitted from the user terminal 20.

The video storage unit 102 is, for example, a recording medium such as a hard disk. The video storage unit 102 stores the acquired video.

The extraction unit 103 is realized, for example, by the operation of a CPU. The extraction unit 103 extracts feature quantities of the applicant U contained in the video. For example, the extraction unit 103 extracts at least one feature quantity of the applicant U's clothing, facial expression, tone of voice, ambient lighting, room background, facial features, body type, hairstyle, and shooting environment contained in the video. The extraction unit 103 also labels the applicant U based on the extracted feature quantities.

The label storage unit 104 is, for example, a recording medium such as a hard disk. The label storage unit 104 stores, for example, labels included in video images that include past applicant U as a subject, and information regarding the pass/fail judgment of past applicant U.

The feature storage unit 105 is, for example, a recording medium such as a hard disk. The feature storage unit 105 stores feature values (labels) that are emphasized depending on the business operator. The feature storage unit 105 stores, for example, the amount of weighting for the feature value (label). For example, if emphasis is placed on the tone of voice feature, the feature storage unit 105 stores a weighting amount that gives more weight to the tone of voice feature.

The recommendation level determination unit 106 is realized, for example, by the operation of a CPU. The recommendation level determination unit 106 determines the recommendation level for applicant U to pass based on the feature values of applicant U contained in the video, video images that include other applicants U as subjects, and information regarding the pass/fail decisions of past applicants U. Specifically, the recommendation level determination unit 106 determines the recommendation level for applicant U to pass based on the labels attached to applicant U, labels attached to past applicants U, and information regarding the pass/fail decisions of past applicants U. In other words, the recommendation level determination unit 106 determines the recommendation level for applicant U to pass based on the labels attached to the video images of past successful applicants. The recommendation level determination unit 106 also determines the recommendation level for applicant U to pass based on feature values related to at least one of the logic of the successful applicant's responses, speaking style, behavior, statements, and the sincerity of the statements. The recommendation level determination unit 106 determines the pass recommendation level using, for example, attractiveness, clarity, fluency, and smile level as feature quantities, as shown in FIG. 3. The recommendation level determination unit 106 displays the pass recommendation level as a percentage. The recommendation level determination unit 106 also determines the pass recommendation level using feature quantities adjusted by the adjustment unit 109, which will be described later.

The judgment result storage unit 107 is, for example, a recording medium such as a hard disk. The judgment result storage unit 107 stores the pass recommendation level determined by the recommendation level determination unit 106 as the judgment result. The judgment result storage unit 107 stores, for example, applicant U and the corresponding pass recommendation level as the judgment result.

The evaluation unit 108 is realized, for example, by the operation of a CPU. The evaluation unit 108 evaluates the trend of the judgment results regarding the degree of recommendation for passing. For example, the evaluation unit 108 evaluates the trend of whether the judgment results for multiple applicants U include many applicants U with high or low specific characteristic amounts. Specifically, the evaluation unit 108 evaluates the trend of whether the judgment results for multiple applicants U include many applicants U with high attractiveness levels. For example, as shown in FIG. 4, the evaluation unit 108 evaluates the trend of the characteristic amounts of all applicants U included in the judgment results.

The adjustment unit 109 is realized, for example, by the operation of a CPU. The adjustment unit 109 adjusts the feature quantities to be emphasized based on the evaluation results. The adjustment unit 109, for example, performs adjustments to increase or decrease the weighting of specific feature quantities. The adjustment unit 109, for example, externally acquires the feature quantities to be adjusted and the adjustment amounts for the feature quantities included in the screen image shown in FIG. 4. The adjustment unit 109 adjusts the weighting based on the acquired feature quantities and adjustment amounts. For example, when it is desired to increase the pass recommendation level for applicant U who has strengths in attractiveness, the adjustment unit 109 performs adjustments to increase the weighting for attractiveness.

The determination unit 110 is realized, for example, by the operation of a CPU. The determination unit 110 determines whether an applicant U who has been judged to have a predetermined level of recommendation for passing has exceeded the pass line. For example, the determination unit 110 determines that an applicant U who meets a predetermined pass line has exceeded the pass line. For example, the determination unit 110 determines whether an applicant U has exceeded the pass line based on a pass line that is predetermined according to the number of people to be passed.

The fulfillment level calculation unit 111 is realized, for example, by the operation of a CPU. The determination unit 110 calculates the fulfillment level for the label for applicant U. The fulfillment level calculation unit 111 calculates the fulfillment level for the label assigned to each applicant U, for example, as shown in FIG. 5.

The analysis unit 112 is realized, for example, by the operation of a CPU. The analysis unit 112 analyzes the trends of applicant U who has been determined to have exceeded the pass line. For example, the analysis unit 112 analyzes the trends in the feature values of applicant U who has exceeded the pass line.

The output unit 113 is realized, for example, by the operation of a CPU. The output unit 113 outputs the degree of recommendation for passing applicant U based on the judgment result. The output unit 113 also outputs the evaluation result by the evaluation unit 108, the degree of satisfaction calculated by the satisfaction calculation unit 111, and the tendency of the feature values analyzed by the analysis unit 112. The output unit 113 outputs, for example, to a display device (not shown) or the like.

Next, the operation of the selection device 1 will be described with reference to the flowchart of FIG.
First, the video acquisition unit 101 acquires a video from the user terminal 20 (step S1). Next, the extraction unit 103 extracts features contained in the video (step S2). Next, the recommendation level calculation unit calculates the pass recommendation level for each applicant U (step S3).

Next, a determination is made as to whether or not the pass line needs to be determined (Step S4). If a pass line determination is required (Step S4: YES), processing proceeds to Step S5. On the other hand, if a pass line determination is not required (Step S4: NO), processing proceeds to Step S7.

In step S5, the judgment unit 110 determines whether applicant U exceeds the predetermined passing grade. Next, the satisfaction level calculation unit 111 calculates the satisfaction level (step S6). Next, processing proceeds to step S7.

In step S7, it is determined whether feature adjustment is necessary. If feature adjustment is necessary (step S7: YES), processing proceeds to step S8. On the other hand, if feature adjustment is not necessary (step S7: NO), processing proceeds to step S10.

In step S8, the evaluation unit 108 evaluates the pass recommendation level judgment result. Next, the adjustment unit 109 adjusts the weighting of the feature amounts based on the adjustment amount obtained based on the judgment result (step S9). Next, processing proceeds to step S10.

In step S10, the output unit 113 outputs the judgment result. The output unit 113 also outputs the judgment result and evaluation result as necessary. This completes the processing of this flow.

Next, the program will be described.
Each component included in the selection device 1 can be realized by hardware, software, or a combination of these. Here, being realized by software means being realized by a computer reading and executing a program.

The program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-RWs, and semiconductor memory (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The display program may also be supplied to a computer via various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The transitory computer-readable media can be supplied to a computer via wired communication paths such as electrical wires and optical fibers, or wireless communication paths.

As described above, the selection device 1 and the program according to this embodiment have the following advantages.
(1) A selection device 1 that selects an applicant U for a specified business based on a video image containing the applicant U as a subject, and includes a video image acquisition unit 101 that acquires the video image, a recommendation level judgment unit 106 that judges the degree of recommendation for passing the applicant U based on the features of the applicant U contained in the video image, video images containing other applicants U as subjects, and information regarding the pass/fail judgment of the other applicants U, and an output unit 113 that outputs the degree of recommendation for passing the applicant U based on the judgment result.
The program also causes a computer to function as a selection device 1 that selects applicant U for a specified business based on video images that include applicant U as a subject, and causes the computer to function as a video image acquisition unit 101 that acquires video images, a recommendation level judgment unit 106 that judges the degree of recommendation for applicant U to pass based on the features of applicant U included in the video images, video images that include other applicants U as subjects, and information regarding the pass/fail judgments of the other applicants U, and an output unit 113 that outputs the degree of recommendation for applicant U to pass based on the judgment results.
This allows the degree of recommendation for passing the applicant U to be output according to the wishes of the business operator, making it easier to decide on successful applicants and determining which applicant U best suits the wishes of the business operator.

(2) The selection device 1 further includes an extraction unit 103 that extracts features of applicant U contained in the video. This makes it possible to easily obtain features used for judgment from the video.

(3) The extraction unit 103 labels the applicant U based on the extracted features, and the recommendation level determination unit 106 determines the pass recommendation level for the applicant U based on the label assigned to the applicant U, the labels assigned to other applicants U, and information regarding the pass/fail decisions of the other applicants U. This makes it possible to determine the applicant U who better matches the operator's preferences based on the other applicants U.

(4) The selection device 1 further includes an evaluation unit 108 that evaluates the tendency of the judgment results regarding the pass recommendation level, and an adjustment unit 109 that adjusts the feature amount to be emphasized based on the evaluation results, and the recommendation level judgment unit 106 judges the pass recommendation level using the adjusted feature amount. This makes it possible to more appropriately adjust the pass recommendation level even if the pass recommendation level is biased.

(5) The recommendation level determination unit 106 determines the recommendation level for applicant U based on the labels attached to the video images of past successful applicants. This makes it possible to determine applicant U who better matches the operator's preferences.

(6) The selection device 1 further includes a judgment unit 110 that determines whether an applicant U who has been judged to have a predetermined or higher pass recommendation level has exceeded the pass line. This allows for a pass/fail judgment to be made, making the selection process easier.

(7) The selection device 1 further includes a satisfaction level calculation unit 111 that calculates the satisfaction level of applicant U with respect to the label. This makes it easy to obtain the reason for the judgment.

(8) The selection device 1 further includes an analysis unit 112 that analyzes the trends of the selected successful applicants. This makes it easy to determine whether the applicant actually meets the operator's expectations.

The above describes preferred embodiments of the selection device and program of the present invention, but the present disclosure is not limited to the above-described embodiments and can be modified as appropriate.

For example, in the above embodiment, the recommendation level determination unit 106 determined the pass recommendation level based on past applicant U, but this is not limited to this. The recommendation level determination unit 106 may also determine the pass recommendation level for actual applicant U by using another applicant U as the ideal image of applicant U envisioned by the business operator.

In addition, in the above embodiment, the selection device 1 is described as including the extraction unit 103, but this is not limited to this. Instead of the extraction unit 103, feature amounts already extracted from video may be acquired. The recommendation level determination unit 106 may use the extracted feature amounts to calculate the pass recommendation level.

In addition, in the above embodiment, the recommendation level determination unit 106 may further determine the pass recommendation level based on information regarding applicant U's status and external evaluation factors. The recommendation level determination unit 106 may determine the pass recommendation level based on information regarding applicant U's career history and test results, for example. The recommendation level determination unit 106 may also use the results of a dark color test as external evaluation factors.

In addition, in the above embodiment, the selection device 1 may include a simulation execution unit 114 that simulates a pass/fail decision to a specified business using the acquired video. The output unit 113 may output the simulation results. For example, the simulation execution unit 114 may simulate a pass/fail decision before actually applying to a business. The output unit 113 may transmit the simulation results to the user terminal 20 that transmitted the video.

In addition, in the above embodiment, the selection device 1 may acquire video images of applicant U for each of multiple businesses. The selection device 1 may also determine the degree of recommendation for passing for each of multiple businesses. The selection device 1 may acquire the video images from the user terminal 20 and also acquire the names of the businesses to which the applicant wishes to apply.

Furthermore, in the above embodiment, labels are not limited to those assigned based on features extracted from videos. For example, labels may be created by creating correct labels based on pre-prepared standards and multiple videos, and then adjusting and determining the labels using the difference between the correct labels and the standards used when judging multiple other videos.

In addition, in the above embodiment, the output unit 113 outputs the pass recommendation level for applicant U, but this is not limited to this. The output unit 113 may output the feature amount without outputting the pass recommendation level. Furthermore, the output unit 113 may output both the pass recommendation level and the feature amount.

In addition, in the above embodiment, the extraction unit 103 may extract attributes of applicant U (such as athletics or international student) from the content included in the video. The recommendation level determination unit 106 may determine the recommendation level for passing based on the attributes.

In addition, in the above embodiment, the simulation execution unit 114 may make a judgment by combining information other than video images (meta information (information on proficiency tests, academic background, etc.)). Furthermore, the simulation execution unit 114 may not be limited to making a binary judgment as to whether the result is a pass or fail, but may instead classify the result into three or more values. Furthermore, the simulation execution unit 114 may make a judgment as to whether the result is a pass or fail using a continuous value, such as a pass rate.

In addition, in the above embodiment, the analysis unit 112 may analyze the tendencies of applicant U who has been determined not to have passed the pass mark.

Furthermore, in the above embodiment, the selection device 1 is not limited to the document screening stage before the first interview, but may be used at any stage in the selection process.

REFERENCE SIGNS LIST 1 Selection device 101 Video image acquisition unit 102 Video image storage unit 103 Extraction unit 106 Recommendation level determination unit 108 Evaluation unit 109 Adjustment unit 110 Judgment unit 111 Satisfaction level calculation unit 112 Analysis unit 113 Output unit 114 Simulation execution unit U Applicant

Claims (14)

A selection device that selects applicants for a predetermined business based on a moving image including the applicants as subjects,
a moving image acquisition unit that acquires the moving image;
a recommendation level determination unit that determines the pass recommendation level of the applicant in accordance with criteria based on multiple feature amounts of the applicant contained in the video, video images including other applicants as subjects, and information regarding the pass/fail decisions of the other applicants;
an evaluation unit that evaluates the tendency of specific features common to a group of applicants with a high pass recommendation level based on the pass recommendation level judgment results for a plurality of applicants ;
an adjustment unit that adjusts a feature quantity to be emphasized among the plurality of feature quantities in the judgment criterion based on an evaluation result, and updates the judgment criterion based on the adjustment result ;
A selection device comprising: The selection device of claim 1 further comprises an extraction unit that extracts the plurality of features of the applicant contained in the video. The selection device described in claim 2, wherein the extraction unit extracts the plurality of features from the video, including at least one feature of the applicant's clothing, facial expression, tone of voice, ambient lighting, room background, facial features, body type, hairstyle, and shooting environment. the extraction unit performs labeling on the applicant based on the extracted plurality of feature amounts;
The selection device described in claim 2 or 3, wherein the recommendation level determination unit determines the recommendation level for passing the applicant based on criteria based on the label attached to the applicant, the labels attached to the other applicants, and information regarding the pass/fail decisions of the other applicants. The selection device described in claim 4, wherein the recommendation level determination unit further determines the applicant's recommendation level for passing based on the criteria adjusted based on labels attached to video images of past successful applicants. The selection device described in claim 5, wherein the recommendation level determination unit further determines the recommendation level for passing the applicant based on feature quantities related to at least one of the logic of the successful applicant's responses, speaking style, behavior, statements, and the sincerity of the statements. A selection device according to any one of claims 3 to 6, wherein the recommendation level determination unit further determines the applicant's recommendation level for passing based on information regarding the applicant's status and external evaluation factors. The selection device described in any one of claims 3 to 7 further comprises a determination unit that determines whether an applicant who has been determined to have a predetermined or higher pass recommendation level exceeds the pass line. The selection device of claim 4 further comprises a fulfillment calculation unit that calculates the fulfillment level of the label for the applicant. The selection device according to claim 2 , further comprising a simulation execution unit that uses the acquired moving image to simulate a pass/fail decision for the predetermined business operator and outputs a simulation result . The selection device of claim 8 further comprises an analysis unit that analyzes the tendencies of applicants who have been determined to have exceeded the pass line. The selection device of claim 8 further comprises an analysis unit that analyzes the tendencies of applicants who are determined not to have passed the pass line. A selection device according to any one of claims 2 to 12, wherein the extraction unit sorts the acquired videos using a predetermined feature from among the plurality of feature amounts. A program that causes a computer to function as a selection device that selects applicants for a predetermined business based on a video image that includes the applicants as subjects,
The computer
a moving image acquisition unit that acquires the moving image;
a recommendation level determination unit that determines the pass recommendation level of the applicant in accordance with criteria based on multiple feature amounts of the applicant contained in the video, video images including other applicants as subjects, and information regarding the pass/fail decisions of the other applicants;
an evaluation unit that evaluates the tendency of specific features common to a group of applicants with a high pass recommendation level based on the pass recommendation level judgment results for a plurality of applicants ;
an adjustment unit that adjusts a feature quantity to be emphasized among the plurality of feature quantities in the judgment criterion based on an evaluation result, and updates the judgment criterion based on the adjustment result ;
A program that functions as a