Utterance situation providing system

Abstract

To provide an utterance situation providing system enabling a user to check a situation in which an utterance occurs when the user examines whether or not the utterance corresponds to an utterance subjected to a predetermined situation.SOLUTION: An utterance situation providing system 100 includes an utterance obtaining device 110 and an utterance situation providing device 130. The utterance obtaining device 110 is placed at a predetermined site, and obtains, as voice data, an utterance at the placed site. The utterance situation providing device 130 analyzes the utterance (the obtained utterance) obtained by the utterance obtaining device 110, determines whether there is a possibility that the obtained utterance corresponds to an utterance (a target utterance) constituting power harassment at a workplace, and notifies a user of the occurrence of the target utterance when it is determined that there is a possibility that the obtained utterance corresponds to the target utterance. Moreover, the utterance situation providing device 130 provides the user with the situation in which the obtained utterance occurs when the user examines whether or not the obtained utterance corresponds to the target utterance.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a speech status providing system, and more particularly to a system that allows a user to check the situation in which a speech occurred when verifying whether or not the speech corresponds to a speech that is subject to a predetermined situation. [Background technology]

[0002] A conventional speech status providing system will be described using a harassment detection system 10 shown in Fig. 11. As shown in Fig. 11, harassment detection system 10 includes multiple devices 12, a server 14, a time-stamping system 16, and a terminal 18, all of which are connected via a network 4 such as the Internet. Server 14 is an example of a harassment detection device.

[0003] The device 12 is a mobile device such as a smartphone that each employee carries for work. The device 12 has a positioning function for acquiring location information, a voice recording function, and a function for measuring physical information data such as pulse. The device 12 transmits audio data, which is constantly recorded during work, and the measured physical information data to the server 14 in association with the employee ID of the owner of the device 12 and the acquired location information data. The audio data is used to detect suspicion of the perpetrator, and the physical information is used to detect suspicion of the victim. Note that a sensor for measuring physical information may be attached to the employee separately from the device 12, and the physical information measured by the sensor may be transmitted to the device 12 via Bluetooth (registered trademark) or the like, and then transmitted from the device 12 to the server 14. The audio data is an example of voice data.

[0004] The server 14 detects the occurrence of harassment based on the recorded data and physical information transmitted from the device 12. By detecting suspicion and the occurrence of harassment in real time by the server 14, it becomes possible to promptly deal with the problem of harassment.

[0005] The time-stamp system 16 is a system for managing the clock-in and clock-out of employees, and transmits the attendance status of employees to the server 14 in response to the employees' clock-in and clock-out.

[0006] Terminal 18 plays back the recorded data acquired at the time when the suspected harassment was detected. Terminal 18 also inputs a determination as to whether the detected harassment is actually harassment or not, and the reason why the harassment occurred, and transmits the result of the harassment determination. Terminal 18 can be realized by a notebook PC, a tablet terminal, a smartphone, etc.

[0007] 12, the server 14 functionally includes a communication unit 20, a database 22, a monitoring control unit 24, a first suspicion detection unit 26, a second suspicion detection unit 28, a harassment detection unit 30, and a harassment determination unit 32. Note that the communication unit 20 is an example of an acquisition unit, the first suspicion detection unit 26 and the second suspicion detection unit 28 are examples of a detection unit, and the harassment detection unit 30 is an example of a detection unit.

[0008] The first suspicion detection unit 26 detects a first suspicion of harassment by the perpetrator based on the audio recording data acquired from the device 12. The perpetrator is an example of a first person. For example, by referring to the audio recording data in the voice physical information record table 48A on that day, if the voice volume of the audio recording data is equal to or exceeds the reference value in the voice volume violation reference table 42A or if the voice recognition result of the audio recording data includes an NG word in the NG word reference table 44A, the first suspicion is detected as having occurred as a violation. Based on the detection result, the first suspicion detection unit 26 updates the suspected perpetrator ID, suspected perpetrator ID (employee ID), suspected perpetrator category, suspected perpetrator content, occurrence time, location information data at the time of occurrence, and audio recording data before and after occurrence in the suspected perpetrator occurrence record table 50A.

[0009] The second suspicion detection unit 28 detects a second suspicion of harassment by the victim based on the physical information data acquired from the device 12. The victim is an example of a second person. For example, by referring to the physical information data in the current day audio physical information record table 48A, if either the heart rate or pulse rate is equal to or greater than the reference value of each physical information data in the stress criteria table 46A, the second suspicion detection unit 28 detects the second suspicion as being stressed. Based on the detection result, the second suspicion detection unit 28 updates the suspected victim ID, suspected victim ID (employee ID), suspected victim category, suspected victim content, time of occurrence, location information data at the time of occurrence, and audio recording data before and after the occurrence in the suspected victim occurrence record table 52A.

[0010] The harassment detection unit 30 detects the occurrence of harassment between the perpetrator and the victim based on the result of matching between the detection time of the first suspicion and the detection time of the second suspicion, and the result of matching between the matching information acquired about the perpetrator and the matching information acquired about the victim. The result of matching between the detection time of the first suspicion and the detection time of the second suspicion is, for example, checked to see if the occurrence times in the suspected perpetrator occurrence record table 50A and the suspected victim occurrence record table 52A partially match for a period of several minutes or the like. The result of matching between the matching information acquired about the perpetrator and the matching information acquired about the victim is, for example, checked to see if the location information data at the time of occurrence in the suspected perpetrator occurrence record table 50A and the suspected victim occurrence record table 52A is within a predetermined range, for example within 10 meters. In addition, the recording data before and after the occurrence of the first suspicion in the suspected perpetrator occurrence record table 50A is similar to the recording data before and after the occurrence of the second suspicion in the suspected victim occurrence record table 52A using a general voice matching technique. If, as a result of these comparisons, for example, the occurrence times are partially identical, the location information data at the time of occurrence is within a specified range, and the recording data before and after the occurrence are similar, harassment detection unit 30 detects that harassment has occurred between the perpetrator and victim. When harassment detection unit 30 detects the occurrence of harassment, it updates harassment suspicion occurrence record notification table 54A and notifies devices 12 owned by the perpetrator and victim that harassment has occurred.

[0011] Harassment determination unit 32 transmits the first recorded data acquired about the perpetrator at the time the first suspicion was detected and the second recorded data acquired about the second person at the time the second suspicion was detected from communication unit 20 to terminal 18 for playback on terminal 18. Note that it is also possible to play only one of the recorded data. Harassment determination unit 32 also accepts input of the harassment determination result from terminal 18 and updates harassment determination table 56A (see Patent Document 1 for the above). [Prior art documents] [Patent documents]

[0012] [Patent Document 1] JP 2020-009238 A Summary of the Invention [Problem to be solved by the invention]

[0013] The above-mentioned harassment detection system 10 has the following points that need to be improved. In the harassment detection system 10, the system determines whether harassment has occurred. However, the occurrence of harassment generally involves a variety of conditions and complex influences, such as the parties involved, the relationship between the parties at the time, the situation at the time, and the process leading up to that situation. Therefore, there is an area that needs to be improved, in that it is difficult to automatically determine whether harassment has occurred based solely on physical changes at the time or emotional expressions such as voice volume.

[0014] Furthermore, the harassment detection system 10 requires employees to carry device 12, such as a smartphone, with them at all times and to keep certain functions running. Therefore, if an employee does not carry device 12 or does not run its functions, the harassment detection system 10 itself will not function. In other words, implementing the harassment detection system 10 places a heavy burden on employees, which is an area that needs improvement.

[0015] The harassment detection system 10 determines whether harassment is suspected by taking into consideration changes in the physical condition of the harasser and the victim. However, there is a problem in that harassment cannot be detected if no change in physical condition occurs at the time of harassment, or if the change is so minor that it cannot be detected.

[0016] Furthermore, harassment detection system 10 requires that both the perpetrator and victim of harassment obtain information to determine whether harassment is suspected, and then compares the two, which complicates processing, and this is an area that needs improvement.

[0017] Furthermore, generally, even though the perpetrator and victim of harassment are in the same place at the time the harassment occurs, the perpetrator and victim are each judged separately, which complicates the process and is an area that needs improvement.

[0018] Therefore, an object of the present invention is to provide a statement status providing system that enables a user to check the situation in which a statement occurred when verifying whether the statement corresponds to a statement that is subject to a specified situation. Effect of the Invention

[0019] The means for solving the problems in the present invention and the effects of the invention are described below.

[0020] The statement status providing system of the present invention is a statement status confirmation system that can confirm the situation in which a statement occurred when a statement that may correspond to a specified situation occurs, and includes a statement acquisition unit that acquires the statement, a statement analysis unit that analyzes the acquired statement acquired by the statement acquisition unit and determines whether the acquired statement may correspond to a target statement that is a statement that is the subject of the specified situation, and stores and retains the acquired statement, the time when the acquired statement was acquired, and the judgment of whether the acquired statement may correspond to the target statement in association with each other, and a statement status providing unit that provides the statement included in a specified period before and / or after the statement that may correspond to the target statement that is the statement that is the subject of the specified situation.

[0021] This allows the user to check the circumstances under which the acquired comment occurred when verifying whether the acquired comment corresponds to the target comment, and thus to determine whether the acquired comment corresponds to the target comment after taking those circumstances into account.

[0022] In the statement status providing system of the present invention, when the statement analysis unit determines that the acquired statement may correspond to a target statement, which is a statement that is subject to the specified situation, it notifies a specified terminal of the occurrence of the acquired statement that may correspond to a target statement, which is a statement that is subject to the specified situation.

[0023] This allows the user to quickly become aware of the occurrence of a utterance that may correspond to the target utterance.

[0024] In the speech status providing system of the present invention, the voice acquisition unit acquires the speech as voice, and the speech status providing unit performs voice recognition on the acquired speech, acquires the text of the acquired speech, and analyzes the speech in the acquired text.

[0025] This makes it possible to easily determine, on a text basis, whether or not an acquired comment corresponds to a target comment.

[0026] In the speech status providing system of the present invention, the speech status providing unit is characterized in that it determines whether the acquired speech is likely to correspond to the target speech using target speech term information having target terms which are terms that, when included in a speech, are judged to be likely to correspond to the specified situation.

[0027] In this way, by storing in advance terms that can be determined to have a high probability of corresponding to a specified situation, it is possible to easily and quickly determine the occurrence of a utterance that may correspond to the target utterance.

[0028] In the speech status providing system of the present invention, the speech status providing unit is further characterized in that it calculates the polarity value of the acquired speech using term polarity information in which a polarity value indicating the degree to which a specified term corresponds to the specified situation is set for the specified term, and stores and retains the calculated polarity value in association with the acquired speech.

[0029] This makes it possible to easily grasp the degree to which the situations in which other statements occurred before and after a statement, including a statement that is determined to have a high probability of corresponding to a specified situation, correspond to a specified situation.

[0030] The statement status providing device of the present invention is a statement status providing device that can confirm the situation in which a statement occurred when a statement that may correspond to a specified situation occurs, and has a statement analysis unit that analyzes an acquired statement, which is an acquired statement, and determines whether or not the acquired statement may correspond to a target statement, which is a statement that is the subject of the specified situation, and stores and retains the acquired statement, the time when the acquired statement was acquired, and the judgment of whether or not the acquired statement may correspond to the target statement in association with each other, and a statement status providing unit that provides the statement included in a specified period before and / or after the statement that may correspond to the target statement, which is a statement that is the subject of the specified situation.

[0031] This allows the user to check the circumstances under which the acquired comment occurred when verifying whether the acquired comment corresponds to the target comment, and thus to determine whether the acquired comment corresponds to the target comment after taking those circumstances into account.

[0032] The statement status providing program of the present invention causes a computer to function as a statement status providing device that can check the situation in which a statement that may correspond to a specified situation occurs when the statement occurs, analyzing an acquired statement, which is an acquired statement, and determining whether or not the acquired statement may correspond to a target statement that is a statement that is the subject of the specified situation, and storing and retaining the acquired statement, the time when the acquired statement was acquired, and the judgment of whether or not the acquired statement may correspond to the target statement in association with each other, and a statement status providing unit that provides the statement included in a specified period before and / or after the statement that may correspond to the target statement that is the statement that is the subject of the specified situation.

[0033] This allows the user to check the circumstances under which the acquired comment occurred when verifying whether the acquired comment corresponds to the target comment, and thus to determine whether the acquired comment corresponds to the target comment after taking those circumstances into account. [Brief description of the drawings]

[0034] [Figure 1] FIG. 1 is a diagram showing an overview of a comment status confirmation system 100 which is an embodiment of a comment status confirmation system according to the present invention. [Diagram 2] FIG. 2 is a diagram showing a hardware configuration of a statement acquisition device 110. [Diagram 3] FIG. 2 is a diagram showing a hardware configuration of a speech status confirmation device 130. [Figure 4] FIG. 13 is a diagram showing the data structure of a comment status information DB. [Diagram 5] FIG. 13 is a diagram showing a data structure of target utterance term information. [Figure 6] FIG. 13 is a diagram showing a data structure of term polarity information. [Figure 7]13 is a flowchart showing a voice acquisition process. [Figure 8] 13 is a flowchart showing an attention statement generation process. [Figure 9] 13 is a flowchart showing a speech status confirmation process. [Figure 10] FIG. 13 is a diagram showing an example of a speech status confirmation screen. [Figure 11] FIG. 1 is a diagram showing a conventional speech status confirmation system. [Figure 12] FIG. 1 is a diagram showing a conventional speech status confirmation system. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0035] The speech status confirmation system according to the present invention will be described in detail below with reference to the drawings. The speech status confirmation system is a system that can confirm the situation in which a speech that may correspond to a predetermined situation occurs when the speech occurs. EXAMPLES

[0036] 1. Speech status providing system 100 The speech status providing system according to the present invention will be described by taking as an example the speech status providing system 100 shown in Fig. 1. The speech status providing system 100 is a speech status checking system that can check the circumstances under which a speech that may be equivalent to power harassment in the workplace occurs. Note that power harassment in the workplace refers to verbal or physical actions that are based on superior-subordinate relationships in the workplace and that harm the working environment of workers by exceeding the scope of what is necessary and reasonable for the business.

[0037] The statement status providing system 100 includes a statement acquiring device 110 and a statement status providing device 130. The statement acquiring device 110 is installed at a predetermined location and acquires statements made at the location as audio. The statement status providing device 130 analyzes the statements acquired by the statement acquiring device 110 (acquired statements), determines whether or not the acquired statements may correspond to statements that are subject to power harassment in the workplace (target statements), and notifies the user of the occurrence of the target statement when it determines that the acquired statements may correspond to the target statements. In addition, the statement status providing device 130 provides the user with the situation in which the acquired statement occurred when the user verifies whether or not the acquired statement corresponds to the target statement.

[0038] This allows the user to quickly learn of the occurrence of a statement that may correspond to the target statement. In addition, when verifying whether the acquired statement corresponds to the target statement, the user can check the circumstances under which the acquired statement occurred, and can therefore determine whether the acquired statement corresponds to the target statement after taking those circumstances into account.

[0039] 2. Hardware Configuration 1. Speech acquisition device 110 The hardware configuration of the utterance acquisition device 110 will be described with reference to FIG. 2. The utterance acquisition device 110 has a control circuit 110a, a microphone 110b, a communication circuit 110c, and a power supply circuit 110d. The control circuit 110a controls the operation of each component of the utterance acquisition device 110. The microphone 110b acquires utterances as voice. The communication circuit 110c transmits and receives various information including voice data acquired through the microphone 110b to and from an external communication device via a predetermined network. The power supply circuit 110d acquires power from an external source and supplies the power required by each component of the utterance acquisition device 110.

[0040] 2. Speech status providing device 130 The hardware configuration of the speech status providing device 130 will be described with reference to Fig. 3. The speech status providing device 130 has a CPU 130a, a memory 130b, a storage 130c such as a hard disk drive (HDD) or a solid state drive (SSD), a keyboard 130d, a mouse 130e, a display 130f, a drive 130g for various media, a communication circuit 130h, and a power circuit 130i.

[0041] The CPU 130a performs processing based on the operating system (OS), the speech status providing program, and other applications recorded in the storage 130c. The memory 130b provides a working area for the CPU 130a. The storage 130c records and holds the operating system (OS), the speech status providing program, and other application programs. The storage 130c also records and holds a speech status information database (hereinafter, speech status information DB), target speech term information, term polarity information, and other information required for the speech status providing program. The speech status information DB, target speech term information, and term polarity information will be described later.

[0042] The keyboard 130d and mouse 130e accept commands from the outside. The display 130f displays images such as a user interface. The drive 130g for various media reads the speech status providing program from a predetermined medium 130p in which the speech status providing program is recorded, and also reads other application programs from other media, thereby reading data from various media. The communication circuit 130h connects to a predetermined network and transmits and receives information to and from external communication devices such as a mobile terminal owned by the user of the speech status providing device 130.

[0043] The power supply circuit 130i obtains power from an external source and supplies the necessary power to each component of the speech status providing device 130.

[0044] 2nd information 1. Speech status information DB The speech status information is information that stores and holds, as text, speech acquired as voice by the voice acquisition device 110. The speech status information DB is information in which multiple pieces of speech status information are accumulated.

[0045] The data structure of the comment status information DB is shown in Figure 4. The comment status information DB has a comment acquisition device identification information description area, an acquisition time description area, a comment identification information description area, a comment text description area, a target comment occurrence description area, and a polarity value description area.

[0046] The statement acquisition device identification information description area describes statement acquisition device identification information that indicates information that uniquely identifies the statement acquisition device. The acquisition time description area describes the acquisition time of the statement (acquired statement) acquired by the statement acquisition device 110 corresponding to the statement acquisition device identification information described in the statement acquisition device identification information description area, that is, the time when the statement corresponding to the acquired statement occurred.

[0047] The speech identification information description area describes speech identification information that identifies the acquired speech. The acquired speech text description area describes the speech corresponding to the acquired speech, that is, text acquired by speech recognition, as acquired speech text information.

[0048] In the target utterance occurrence description area, a flag is written to indicate that the acquired utterance includes a target term described in the target utterance term information and may correspond to a utterance (target utterance) that is the subject of power harassment in the workplace. Specifically, when the acquired utterance text information includes a target term described in the target utterance term information, the flag is written.

[0049] The polarity value description area describes a polarity value that indicates the degree to which the acquired comment corresponds to a specific situation. The polarity value is calculated for each acquired comment by analyzing the acquired comment and using the term polarity information. The term polarity information and the calculation of the polarity value will be described later.

[0050] For example, when acquiring audio indicating the utterance "I'm sure I told you how to do that before" from a utterance acquisition device 110 with identification number "0001" at "August 1, 2023, 14:15:00," as shown in FIG. 4, in the utterance status information DB, the identification number description area of ​​the utterance acquisition device describes the identification number of the utterance acquisition device 110, "0001," the acquisition time description area describes "2023 / 8 / 1 / 14:15:00," the utterance identification information description area describes information that identifies the acquired utterance with a single digit, here "0020," and the text of the utterance acquired as audio describes "I'm sure I told you how to do that before" in the acquired utterance text information description area.

[0051] Furthermore, since the text "I'm sure I've said that about how to do it before" described in the acquired utterance text information contains the target term "I've said that before" described in the target utterance term information, a flag "○" indicating that the target utterance has occurred is described in the corresponding target utterance occurrence description area of ​​the utterance situation information DB. Furthermore, a polarity value calculated using the term polarity information is described for the text "I'm sure I've said that about how to do it before" described in the acquired utterance text information. Note that the description of the polarity value is omitted in FIG. 4.

[0052] 2. Target speech term information The target utterance term information is information for determining whether a utterance acquired by the utterance acquisition device 110 corresponds to a predetermined situation based on whether a predetermined term is included. Note that terms to be stored in the target utterance term information are selected in advance depending on the predetermined situation to be set.

[0053] The data structure of the target utterance term information is shown in Figure 5. The target utterance term information has a target term description area.

[0054] In the target term description area, terms that are determined to be highly likely to correspond to a set predetermined situation when included in a utterance are described as target terms. Note that target terms may be words, phrases, or sentences. For example, in FIG. 5, terms that are determined to be likely to indicate "power harassment" or so-called "power harassment" when included in a utterance are described as target terms. Here, words such as "you" and "idiot" and sentences or phrases such as "watch your language" and "I told you before" are described.

[0055] 3.Term polarity information The term polarity information is set for a given term a polarity value indicating the degree to which the term corresponds to power harassment in the workplace.

[0056] The data structure of the term polarity information is shown in Fig. 6. The term polarity information has a target term description area, a reading description area, a part of speech description area, and a polarity value description area.

[0057] In the target term description area, various commonly used terms are described as target terms. The target terms are described in the form of morphemes used when speech is recognized and the acquired text is morphologically analyzed. Speech acquisition device identification information indicating information for uniquely identifying a speech acquisition device is described. In the reading description area, the reading of the term described in the term description area is described. In the part of speech description area, the part of speech of the term described in the term description area is described. In the polarity value description area, a value indicating how likely the term described in the term description area is to apply to a specified situation is described as a polarity value. The polarity value is set between "+1" and "-1" and closer to "+1" if the possibility of applying to a specified situation is high.

[0058] For example, in the term polarity value information shown in FIG. 6, a situation that corresponds to so-called "power harassment" is set as the predetermined situation, and the term "praise" is set to "-0.999" as the polarity value.

[0059] The term polarity information is used to calculate the degree to which the acquired comment is likely to correspond to a predetermined situation, i.e., the polarity value of the acquired comment. The polarity value of the acquired comment is calculated, for example, by determining whether or not the acquired comment contains each target term described in the term polarity information, and adding up the polarity values ​​that indicate the degree to which the acquired comment corresponds to the predetermined situation, which are set for the predetermined term.

[0060] 2. Operation of speech status confirmation system 100 The comment status confirmation system 100 executes a comment acquisition process by the comment acquisition device 110, and a comment analysis process and a comment status providing process by the comment status providing device 130. The comment acquisition process by the comment acquisition device 110, and the comment analysis process and the comment status providing process by the comment status confirmation device 130 will be described below.

[0061] 1. Comment acquisition process The utterance acquisition process in the utterance acquisition device 110 will be described with reference to the flowchart shown in Fig. 7. When the utterance acquisition device 110 acquires a utterance as audio via the microphone 110b (S701), it generates acquired utterance information that associates acquired utterance audio information, which is audio information of the acquired utterance, with utterance acquisition device identification information that identifies the device itself and acquisition time information, which is the time when the utterance was acquired (S703).

[0062] When utterances are continuously acquired as voice, acquired utterance information is generated based on the occurrence of a predetermined interval, the passage of a predetermined time, or the like.

[0063] The comment acquisition device 110 transmits the generated acquired comment information to the comment status providing device 130 via the communication circuit 110c (S705).

[0064] If the comment acquisition device 110 determines that it cannot transmit the acquired comment information to the comment status confirmation device 130, it temporarily stores the information in a memory within the control circuit 110 and transmits it to the comment status reporting device 130 at a specified time.

[0065] The statement acquisition device 110 repeats the processes of steps S701 to S705 until the operation is completed (S707).

[0066] 2. Speech analysis processing The statement analysis process in the statement status providing device 130 will be described with reference to the flowchart shown in Fig. 8. When the CPU 130a of the statement status providing device 130 acquires acquired statement information from the statement acquiring device 110 (S801), it executes a voice recognition process for converting the acquired statement voice information into text (S803). Note that the voice recognition process uses a voice recognition engine, a voice recognition service, or the like that is generally used and provided. Also, the voice recognition process may use AI (Artificial Intelligence) technology.

[0067] The CPU 130a records and saves the acquired utterance text information, which is text acquired by voice recognition processing of the acquired utterance voice information, in the utterance status information DB in association with the utterance acquisition device identification information and the acquisition time associated with the acquired utterance information (S805).

[0068] The CPU 130a judges whether the acquired utterance text information acquired in step S803 has the target term described in the target utterance term information, that is, whether the acquired utterance may correspond to a utterance that is the subject of power harassment in the workplace (target utterance) (S807). If the CPU 130a judges that the acquired utterance text information has the target term, that is, may correspond to the target utterance, it records a flag indicating that it may correspond to the target utterance in the target utterance occurrence description area corresponding to the acquired utterance text information acquired in step S803 in the utterance status information DB (S809).

[0069] Then, the CPU 130a transmits target utterance occurrence information indicating that a utterance including the target term has been made to a predetermined communication terminal (S811). The target utterance occurrence information displays, for example, the time when the utterance including the target term was made, here, the acquisition time of the acquired utterance, and information identifying the utterance acquisition device that acquired the acquired utterance, for example, utterance acquisition information identification information.

[0070] The CPU 130a calculates the polarity value of the acquired comment (S813). The polarity value for the acquired comment is calculated, for example, by performing a morphological analysis on the acquired comment text information acquired in step S803, determining whether or not each target term described in the term polarity information is included, and using the polarity values ​​corresponding to each target term, for example, by adding them up.

[0071] The CPU 130a records the calculated polarity value in the polarity value description area corresponding to the acquired comment text information acquired in step S803 in the comment status information DB (S815).

[0072] The CPU 130a repeats the processes from step S801 to step S815 until the operation is completed (S817).

[0073] 3. Speech status provision processing The target comment occurrence information acquirer who has acquired the target comment occurrence information can confirm and verify the situation in which the comment occurred via the comment status providing device 130.

[0074] Through the acquired target utterance occurrence information, the person who acquired the target utterance occurrence information can confirm information that identifies the utterance that is the subject of the target utterance occurrence information, such as utterance acquisition device identification information for identifying the location where the utterance that is the subject of the target utterance occurrence information occurred, and the time when the utterance that is the subject of the target utterance occurrence information occurred.

[0075] The target comment occurrence information acquirer uses a specified terminal to transmit comment search screen request information to the comment status confirmation device 130, requesting the transmission of a comment search screen for searching for the comment that is the subject of the target comment occurrence information.

[0076] When the CPU 130a acquires the comment search screen request information (S901), it transmits a predetermined comment search screen (S903).

[0077] The target utterance occurrence information acquirer acquires a utterance search screen, inputs utterance identification information, which is information for identifying the utterance that is the subject of the target utterance occurrence information, via a terminal that displays the screen, and transmits the information to the utterance status confirmation device 130 via a specified terminal. Note that the utterance identification information may be information that identifies the utterance that is the subject of the target utterance occurrence information that can be confirmed by the target utterance occurrence information acquirer in the target utterance occurrence information, such as utterance acquisition device identification information for identifying the location where the utterance that is the subject of the target utterance occurrence information occurred, or the time when the utterance that is the subject of the target utterance occurrence information occurred.

[0078] When the CPU 130a of the comment status checking device 130 acquires comment identification information (S905), it acquires comment status information related to the comment that can be identified by the comment identification information and comment status information included within a predetermined time before and after the comment that can be identified by the comment identification information from the comment status information DB (see FIG. 4) (S907). The CPU 130a generates comment status providing information for providing the acquired comment status information to a predetermined terminal (S909). The CPU 130a transmits the generated comment status providing information to the predetermined terminal (S911).

[0079] An example of a screen displaying speech status information on a specified terminal is shown in Fig. 10. The speech status confirmation screen has a warning display area R110, a polarity value display area R130, a speech text display area R150, and a voice acquisition location display area R170. In the warning display area R110, the presence of the speech that is the subject of the warning is displayed on a time axis in a specified display format C110 for a specified time before and after (hereinafter, "display time") the time when the speech that is the subject of the search warning was made. Note that, in the speech status confirmation screen, the specified display format C110 is displayed in a selectable format.

[0080] In the polarity value display region R113, the polarity value for each message is displayed on the time axis during the display time in a predetermined display form C130. Note that in the message status confirmation screen, the predetermined display form C130 is displayed in a selectable form, similar to the display form C110 in the warning display region R110.

[0081] When the display form C110 in the warning display area R110 or the display form C130 in the polarity value display area R130 is selected, the speech text information of the utterance corresponding to the selected display form C110 or display form C130 is displayed in the utterance text display area R130 for a predetermined time including the time when the utterance was made. Also, the acquired utterance text information C150 related to the utterance corresponding to the selected display form C110 or display form C130 is displayed distinguished from other speech text information.

[0082] When the time range to be displayed is changed, new comment identification information is transmitted to the comment status providing device 130. The new comment identification information includes the comment acquisition device identification information of the comment acquisition device 110, the time of the comment to be displayed, and the like.

[0083] 9, when the CPU 130a of the speech status check device 130 acquires speech status provision end information indicating that the current speech status providing process is to be ended (S913), the CPU 130a ends the current speech status providing process (S915). The current speech status providing process is ended by selecting a predetermined end button displayed on the screen. The CPU 130a repeats the processes of steps S901 to S915 until the operation is ended (S917).

[0084] In this way, the acquirer of the target utterance occurrence information can confirm the occurrence situation of the utterance that is the subject of the target utterance occurrence information via the utterance situation providing information. When a certain type of utterance occurs, the meaning of the utterance cannot be determined from the utterance alone, or judging from the utterance alone may lead to misunderstanding, so there are many cases where it is necessary to confirm the situation in which the utterance occurred, that is, the utterance situation. By using the utterance situation providing system 100 in this embodiment, even if a utterance that corresponds to a predetermined situation occurs, the user can easily confirm the situation in which the utterance occurred and take appropriate measures against the situation in which the utterance occurred.

[0085] Furthermore, the messages are displayed on a timeline based on the polarity values, which allows the user to easily determine whether a series of messages corresponds to a given situation based on the distribution of the message polarity values.

[0086] [Other embodiments] (1) Predetermined situation: In the above-mentioned embodiment 1, "power harassment" was set as the situation for determining whether to issue a warning, but it is not limited to the example, so long as it is a situation in which it is necessary to check the situation of the uttered remark. For example, it may be "sexual harassment", "moral harassment", "customer harassment", etc.

[0087] Furthermore, it does not have to be a negative situation such as "harassment," but rather a positive situation, such as when a compliment is given.

[0088] (2) Speech status confirmation screen: In the above-mentioned embodiment 1, the warning display area R110, the polarity value display area R130, and the speech text display area R150 are displayed on the same screen on the confirmation status confirmation screen, but each area may be displayed as a separate screen and transitions between them may be made possible.

[0089] Furthermore, necessary areas may be arranged on the same screen, such as the warning display area R110 and the comment text display area R150 on the same screen, or the warning display area R110 and the polarity value display area R130 on the same screen.

[0090] (3) Warning Information: In the above-described first embodiment, the warning information is transmitted to a predetermined terminal. However, the warning information may be displayed in a necessary place such as the display 130f of the remark warning device 130 as appropriate.

[0091] (4) Hardware configuration of the remark warning device 130: In the above-mentioned first embodiment, the remark warning device 130 is formed using the CPU 130a, etc. However, as long as it can execute various processes according to the present invention, it is not limited to the illustrated example. For example, various processes may be executed using a dedicated logic circuit.

[0092] (5) Hardware configuration of the voice capturing device 110: In the above-mentioned first embodiment, the voice capturing device 110 is formed as a device separate from the utterance warning device 130 in the voice warning system 100, but is not limited to the illustrated example as long as it can capture voice. For example, a microphone may be connected to the voice warning device 130, and the function of the voice capturing device 110 may be integrated into the utterance warning device 130.

[0093] In addition, although the voice acquisition device 110 transmits the voice to the voice warning device 130 each time it acquires a voice, it may also be possible to store the voice in a specified recording medium and have the user provide the voice information recorded in the recording medium to the voice warning device 130 as appropriate.

[0094] (8) Voice warning program: In the above-mentioned embodiment 1, the voice warning program is described as realizing processing in accordance with the exemplary flowchart, but is not limited to the exemplary one as long as it realizes similar processing. [Industrial Applicability]

[0095] The voice warning system according to the present invention can be used, for example, in workplaces, customer centers, and call centers. [Explanation of symbols]

[0096] 100 Voice Warning System 110 Audio capture device 110a Control circuit 110b Microphone 110c communication circuit 110d power circuit 130 Audio warning device 130a CPU 130b Memory 130c Storage 130d Keyboard 130e Mouse 130f Display 130g Drive for various media 130h Communication circuit 130i power circuit 130p Various media

Claims

1. A speech status confirmation system capable of confirming the situation in which a speech that may correspond to a predetermined situation occurs, comprising: a comment acquisition unit for acquiring the comment; a statement analysis unit that analyzes acquired statements acquired by the statement acquisition unit, judges whether or not the acquired statement is likely to correspond to a target statement that is a statement that is the subject of the specified situation, and stores and holds the acquired statement, the time when the acquired statement was acquired, and the judgment as to whether or not the acquired statement is likely to correspond to the target statement in association with each other; a statement status providing unit that provides statements included before and / or after a predetermined period of time with respect to the statements that may correspond to target statements that are statements that are the subject of the predetermined situation; A speech status providing system having the above configuration.

2. In the speech status providing system according to claim 1, The statement analysis unit when it is determined that the acquired comment may correspond to a target comment that is a comment that is a subject of the predetermined situation, notifying a predetermined terminal of the occurrence of the acquired comment that may correspond to a target comment that is a comment that is a subject of the predetermined situation; A speech status providing system characterized by the above.

3. In the speech status providing system according to claim 2, The voice acquisition unit Acquire the utterance as audio; The speech status providing unit, performing speech recognition on the acquired utterance, acquiring a text of the acquired utterance, and analyzing the utterance in the acquired text; A speech status providing system characterized by the above.

4. In the speech status providing system according to claim 3, The speech status providing unit, determining whether the acquired utterance is likely to correspond to the target utterance by using target utterance term information having a target term which is a term determined to be likely to correspond to the predetermined situation when included in the utterance; A speech status providing system characterized by the above.

5. In the speech status providing system according to claim 4, The comment status providing unit further Calculating a polarity value of the acquired comment using term polarity information in which a polarity value indicating the degree to which a predetermined term corresponds to the predetermined situation is set, and storing and retaining the calculated polarity value in association with the acquired comment; A speech status providing system characterized by the above.

6. A speech status providing device capable of confirming a situation in which a speech that may correspond to a predetermined situation occurs, comprising: a statement analysis unit that analyzes an acquired statement, determines whether or not the acquired statement is likely to correspond to a target statement that is a statement that is the subject of the specified situation, and stores and holds the acquired statement, the time when the acquired statement was acquired, and the determination as to whether or not the acquired statement is likely to correspond to the target statement in association with each other; a statement status providing unit that provides statements included before and / or after a predetermined period of time with respect to the statements that may correspond to target statements that are statements that are the subject of the predetermined situation; A speech status providing device having the above configuration.

7. Computer, A speech status providing device capable of confirming a situation in which a speech that may correspond to a predetermined situation occurs, comprising: a statement analysis unit that analyzes an acquired statement, determines whether or not the acquired statement is likely to correspond to a target statement that is a statement that is the subject of the specified situation, and stores and holds the acquired statement, the time when the acquired statement was acquired, and the determination as to whether or not the acquired statement is likely to correspond to the target statement in association with each other; a statement status providing unit that provides statements included before and / or after a predetermined period of time with respect to the statements that may correspond to target statements that are statements that are the subject of the predetermined situation; A program that provides speech status.

Images