JP7231806B2 - Information processing device, information processing system, control method, and program - Google Patents

Description

The present invention relates to a technique for assisting error proofreading of speech recognition results. It relates to a technology that enhances the accuracy of information security for readers who see the results.

Conventionally, research and development of voice recognition for converting human utterances into character strings have been carried out. As a practical application, it has been used for purposes such as subtitling television broadcasts and deaf people understanding other people's speech.

Especially in recent years, with the progress of machine learning, etc., practical recognition accuracy has come to be realized. However, sufficient recognition accuracy has not yet been achieved, and it is necessary to pay particular attention to how the speaker speaks. For example, recognition accuracy varies greatly depending on the distance between the microphone and the mouth and the clarity of speech.

Therefore, software for correctly correcting the character string of the speech recognition result on a personal computer or the like has been realized.

Originally, there was a service called "computer summary writing" that provided deaf people by listening to the speaker's utterance and inputting it into a computer, but with the advent of software that uses speech recognition technology, it is now possible to support the work of the person who inputs. It is becoming possible.

Such software generally divides the speech at the timing at which the speech is interrupted, and displays the speech recognition results (character strings) on the screen in chronological order in units of the divisions. Those strings will be corrected by a proofreader trained as a computer abstract transcriber.

However, utterances are usually faster than the work of correcting recognition results, that is, the keyboard input work in an information processing apparatus, so the burden on the proofreader who performs the correction work has not yet been sufficiently alleviated.

Patent Literature 1 provides a voice recognition result editing device that assists a proofreader in correcting misrecognition of voice recognition.

In the technique disclosed in Patent Literature 1, an utterance input from a microphone is converted into a character string with a degree of reliability assigned to each word by a speech recognition unit. At this time, not only words with the highest degree of confidence, but also words that satisfy a predetermined condition, such as words with certainty or more, are converted into character strings and included in the set of speech recognition results. The correct recognition result can be selected/corrected from the correct candidates of (Patent Document 1, paragraph 0013, FIG. 8).

JP 2017-040856 A

However, since one utterance includes a plurality of words, with the technique of Patent Document 1, when the number of words is large, the number of recognition results becomes enormous, making it difficult to display on a display device in an easy-to-understand manner. Especially in the case of the personal computer summary writing, since the correction is made in an easy-to-understand manner by judging from the context, the character string of the utterance before the utterance being corrected may be checked, but there is a possibility that there is no such area. In other words, simply displaying candidates for recognition results may not provide efficient assistance to the proofreader.

Also, if there are not enough proofreaders, it may not be possible to correct all speech recognition errors. In this case, it is necessary to determine the part to be corrected preferentially, but with the technique described in Patent Document 1, speech recognition results that have not been corrected for a certain period of time can be corrected regardless of their importance. (timeout processing described in paragraph 0022 of Patent Literature 1, FIG. 5). be.

Moreover, although it is ideal that the proofreader corrects the character strings corresponding to all utterances, in reality, there are cases where it is impossible. If it is not possible, it is necessary to give the person (for example, deaf person) who sees the correction result as easy-to-understand information as possible by having the proofreader make corrections with appropriate priority.

SUMMARY OF THE INVENTION In view of the above problems, an object of the present invention is to display information indicating the order of priority for proofreading for a series of text data, which is the result of recognition of speech data segmented from continuous speech data, in an efficient manner for the user. It is to provide a technique for correcting it.

An information processing device for acquiring a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data, Output for causing a display device to display information indicating a priority order in which the series of text data for each of the segmented voice data should be corrected based on the first certainty of the series of text data. and receiving means for receiving correction of the text data in order to update the text data in which the information is displayed. The display device displays information indicating the order of priority in which the series of text data should be corrected based on the first certainty adjusted based on whether the text data is a part of speech.

In view of the above problems, according to the present invention, information indicating the order of priority to be corrected is displayed for a series of text data, which is the result of recognition of speech data segmented from continuous speech data, so that the user can make corrections efficiently. It is possible to provide technology for

It is a figure which shows an example of the system configuration|structure which concerns on embodiment of this invention. 1 is a block diagram showing an example hardware configuration of an information processing apparatus according to an embodiment of the present invention; FIG. It is a figure showing an example of functional composition concerning an embodiment of the present invention. FIG. 5 is a diagram showing an example of a screen displaying speech recognition results according to the embodiment of the present invention; FIG. 4 is a diagram showing an example of processing from voice input to proofreading distribution according to the embodiment of the present invention; It is a figure which shows an example of the data format of the recognition result which concerns on embodiment of this invention, and the reliability of a recognition result. FIG. 4 is a diagram showing an example of a flowchart for explaining processing up to prioritization for analysis and proofreading of speech recognition results according to the embodiment of the present invention. FIG. 5 is a diagram showing an example of a flowchart for explaining a process of prioritization for proofreading according to an embodiment of the present invention; FIG. 4 is a diagram showing an example of a storage unit that stores information used for priority processing according to the embodiment of the present invention; FIG. 4 is a diagram for showing an example of a result of recalculating the certainty factor of the speech recognition result according to the embodiment of the present invention; FIG. 4 is a diagram showing an example of a user interface that displays speech recognition results according to the embodiment of the present invention; FIG.

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a diagram showing an example of a system configuration according to an embodiment of the invention.
<System configuration example 1>

A system according to the embodiment of the present invention comprises a speech recognition server 101 and an information processing terminal 102 (102a for speaker, 102b for proofreader, and 102c for reader). The user inputs voice with the microphone 104 connected to the information processing terminal 102a. The information processing terminal 102a transmits the voice to the voice recognition server 101, converts it into a character string, sends it to the information processing terminals 102a to 102c, and displays it on the information processing terminals 102a to 102c to present it to the user. That is, the information processing terminals 102a to 102c may have both inputs and outputs for voice input and character string output. The information processing terminal 102 for outputting here may serve both as a reader terminal 102c and a proofreader terminal 102b, which will be described later, or may be dedicated information processing terminals. Also, although the output is performed on the display device connected to the information processing terminal 102, a configuration using a projector or the like is also included in the system configuration related to the embodiment of the present invention. If a projector is used, only one information processing terminal 102 is provided for the speaker, and all readers may read the character string of the voice recognition result displayed on the screen from the projector connected to the information processing terminal 102a. In that case, the speaker himself/herself or another user may directly correct misrecognition as a proofreader on the information processing terminal 102a for the speaker.

Furthermore, the voice recognition server 101 may exist on the cloud, and in that case, the user of this system may use the functions of the voice recognition server 101, which will be described later, as a cloud service. Even if these services are used, the system configuration is related to the embodiment of the present invention.
<System configuration example 2>

The information processing terminals 102a to 102c described in Configuration Example 1 have both input and output functions, but may be divided into input-only terminals and output-only terminals.
<System configuration example 3>

The voice recognition server 101 and the information processing terminals 102a to 102c may be housed in the same housing. That is, one of the information processing terminals 102a to 102c shown in FIG.

FIG. 2 is a block diagram showing an example of a hardware configuration applicable to the voice recognition server 101 and information processing terminals 102a to 102c according to the embodiment of the present invention.

As shown in FIG. 2, the speech recognition server 101 and the information processing terminals 102a to 102c are connected via a system bus 204 to a CPU (Central Processing Unit) 201, a RAM (Random Access Memory) 202, a ROM (Read Only Memory) 203, A configuration in which an input controller 205, a video controller 206, a memory controller 207, a communication I/F controller 208, and the like are connected is adopted. The CPU 201 comprehensively controls each device and controller connected to the system bus 204 .

The ROM 203 or the external memory 211 also stores a BIOS (Basic Input/Output System), which is a control program for the CPU 201, an OS (Operating System), and other data necessary for realizing functions executed by each server or each PC, which will be described later. Various programs are stored. It also stores information necessary for carrying out the present invention. Note that the external memory may be a database.

A RAM 202 functions as a main memory, a work area, and the like for the CPU 201 . The CPU 201 loads necessary programs and the like from the ROM 203 or the external memory 211 to the RAM 202 when executing processing, and implements various operations by executing the loaded programs.

The input controller 205 also controls inputs from a keyboard (KB) 209 and a pointing device such as a mouse (not shown).

Video controller 206 controls display on a display such as display 210 . The display may be a display such as a liquid crystal display. These are used by administrators as needed.

The memory controller 207 is an external storage device (hard disk (HD)), flexible disk (FD), or PCMCIA (Personal Computer Memory Card International Association) Controls access to external memory 211 such as compact flash (registered trademark) memory connected to a card slot via an adapter.

A communication I/F controller 208 connects and communicates with an external device via a network, and executes communication control processing in the network. For example, communication using TCP/IP (Transmission Control Protocol/Internet Protocol) is possible.

It should be noted that the CPU 201 can display on the display 210 by, for example, rasterizing an outline font to a display information area in the RAM 202 . The CPU 201 also allows the user to issue instructions using a mouse cursor (not shown) or the like on the display 210 .

Various programs described later for realizing the present invention are recorded in the external memory 211 and are executed by the CPU 201 by being loaded into the RAM 202 as necessary.
FIG. 3 is a diagram showing an example of a functional configuration according to an embodiment of the invention.

The information processing terminal 102 may have the functions of the speaker 102a, the proofreader 102b, and the reader 103b in separate terminals or in a common terminal. to explain.

The voice acquisition unit 311 inputs the utterance of the speaker's voice as voice data from a microphone built in or connected to the information processing terminal 102 , and transmits the voice data to the voice recognition server 101 by the voice data transmission unit 312 .

The speech recognition server 101 transfers the speech data received by the speech data reception unit 321 to the speech recognition unit 322, converts the speech data into a character string, and sends the recognition result to the information processing terminal 102 by the recognition result transmission unit 323. send back as Further, the recognition result described above is stored in the recognition result storage unit 320 by the recognition result management unit 324 .

The information processing terminal 102 receives the character string at the recognition result receiving unit 313 and presents it to the reader (the user of the information processing terminal 102) by displaying it on the display unit 314. FIG.

The priority determination unit 325 allows a proofreader who uses the information processing terminal 102 to proofread an error in speech recognition to identify a character string that should be proofread preferentially. Give priority to recognition results.

The prioritized character strings are transmitted to the information processing terminal 102, and the display unit 314 of the information processing terminal 102 identifiably displays the priority to be proofread by the proofreader as described above. The recognition result proofreading unit 315 provides a function for proofreading an error in the recognition result by the proofreader editing the character string.

The proofreading result is transmitted to the voice recognition server 101 by the proofreading result transmitting unit 316 of the information processing terminal 102 , received by the proofreading result receiving unit 326 of the voice recognition server 101 , and is stored in the recognition result storage unit 320 . Update results.

The updated recognition result is distributed by the proofreading result distributing unit 327 to the information processing terminals 102 other than the information processing terminal 102 used by the proofreader for proofreading, so that the reader can see the proofreading result. Presented.
FIG. 4 is a diagram showing an example of a screen displaying speech recognition results according to the embodiment of the present invention.

An utterance example 400 is an utterance example of a speaker at a conference, a lecture, or the like. It is not necessary to specify a single speaker. For example, in a meeting, there is a possibility that someone other than the chairperson will speak, and in a lecture, there may be other speakers besides the speaker, such as the moderator and questioners. good too.

In the utterance example 400, it is divided into A to K, and these are divisions of the utterance of the speaker. For example, it indicates a case where there is a blank (silent state) for a certain period of time in the utterance.

On the other hand, the speech recognition result display screen 401 is also partitioned corresponding to the above A to K (a plurality of display frames 404A to K), but these are separated by the speech recognition unit 322 of the speech recognition server 101. It recognizes the state of and separates the character string of the recognition result. These are divided and stored in the recognition result storage unit 320, and the display unit 314 of the information processing terminal 102 divides and displays them in an easy-to-understand manner for the reader. This is an example and does not necessarily delimit 404, it is just a matter of design. It is sufficient that the order of priority for calibration, which will be described later, is displayed in a recognizable manner.

A start button 402 is pressed to allow the speech recognition server 101 to recognize an utterance. The system configuration diagram (FIG. 1) shows a plurality of information processing terminals 102 and microphones connected to them. It is for It is possible to input speech to only one information processing terminal 102, or to input speech to a plurality of information processing terminals 102 at the same time, depending on the design of the system. Also, there may be an end button 403 for notifying the information processing terminal 102 that no utterance has been input in response to the start button 402 .

Of the 404A to 404K, 404A to 404J indicate a state in which the "fixed time blank (silent state)" has passed. On the other hand, 404K is a state in which the speech recognition unit 322 still determines that the speech of the speaker is continuing, indicating that the output of the recognition result is being continued. In the figure, part of the utterance is assumed to have already been recognized and is displayed.

FIG. 5 is a diagram showing an example of processing from voice input to proofreading distribution according to the embodiment of the present invention. Each step of the flowchart of FIG. 5 is executed by the CPU 201 on the speech recognition server 101 and the CPU 201 on the information processing terminals 102a to 102c.

In step S501, the speech of the speaker is received through a microphone or the like connected to the information processing terminal 102a and converted into voice data.

The information processing terminal 102a transmits the voice data to the voice recognition server 101 in step S502, and the voice data is received by the voice recognition server 101 in step S503.

In step S504, the voice recognition server 101 converts the utterance of the speaker in the voice data into a character string by voice recognition. The character string of the recognition result is delimited by utterance units as described above, and further delimited by linguistic units such as morphemes so as to be identifiable. The result of speech recognition is given not only a character string but also a degree of certainty indicating the degree of probability that the speech recognition unit 322 estimates that the recognition result is correct. In addition, when it is divided into linguistic units such as morphemes, each morpheme may be tagged with a degree of certainty and a detailed part of speech. Part-of-speech assignment by morphological analysis will be explained with an example in FIG. It can be classified as finely. Morphological analysis and speech recognition are well-known techniques, and detailed descriptions thereof will be omitted.

In step S<b>505 , the speech recognition server 101 transmits the character string resulting from the conversion in step S<b>504 to the information processing terminal 102 . When a plurality of information processing terminals 102 are connected in the system, the character string is transmitted not only to the information processing terminal 102a to which the utterance is input, but to all the information processing terminals 102 as well. The speech recognition result may also be transmitted to the information processing terminal 102a used by the speaker to input the speech data so that the speaker himself/herself can check the speech recognition result. The information processing terminal 102 receives the character string in step S506.

In step S<b>507 , the speech recognition server 101 stores the speech recognition result in the recognition result storage unit 320 . The format in which the recognition results are stored will be described in detail with reference to FIG.

FIG. 6 is a diagram showing an example of the data format of the recognition result and the confidence factor of the recognition result according to the embodiment of the present invention. As an example, it is assumed that the recognition result is stored in the structure of the recognition result information 600 .

601A-J are data corresponding to the utterances A-J in FIG. The character string resulting from the speech recognition unit 322 is stored in the recognized character string 603 corresponding to the break of speech as described above. 602A-J are confidence factors corresponding to each of utterances 601A-J. Each recognition result is composed of a recognized character string 603, a morpheme notation 604 to be described later, and a certainty 605 of the recognition result of each morpheme.

In addition, morphological notation 604 with a dark background (such as 606) will be described later using FIG. 8 to FIG. used to determine which part to calibrate.

In step S508, the speech recognition server 101 manages one or more pieces of data newly input by the utterance and stored in the recognition result storage unit 320 in the processes up to step S507, determines the priority of proofreading, and other management. That is, it manages the recognition result information 600 of FIG. These processes will be described in detail with reference to FIGS. 7 and 8. FIG.

Asynchronously with the processing in the speech recognition server 101, the information processing terminal 102b for the proofreader presents the character string received in step S506 to the proofreader on the display device of the information processing terminal 102b, and in step S509, Receive proofreading work from proofreaders. The proofreading work of the proofreader is to edit the character string corresponding to the utterance while following the identifiable priority displayed on the display device of the information processing terminal 102b. The screen during calibration will be described later with reference to FIG. Also, in step S509, when the proofreading work is started, the fact is notified to the speech recognition server 101, and the correction state of the data stored in the recognition result storage unit 320 is changed to "under proofreading".

In step S510, the character string as a result of the proofreading described above is transmitted from the information processing terminal 102b. Update data. At this time, the correction status is changed to "completed" and the necessity of correction is changed to "unnecessary".

In step S<b>512 , the proofreading result distributing unit 327 of the speech recognition server 101 transmits to the information processing terminal 102 the proofread-completed character string, that is, the character string in which the erroneously recognized portion in the speech recognition has been proofread.

The information processing terminal 102b for the proofreader who corrected the error already displays the correct character string at the time of proofreading, but as a design matter, the correct character string is also sent to the information processing terminal 102b, that is, to itself. You may In addition, in the flowchart of FIG. 5, the character string for which proofreading has been completed is once distributed to the information processing terminal 102 via the speech recognition server 101, but the information processing terminal 102b for proofreading directly distributes it to other information processing. It may be distributed to terminals 102 . This difference is only a matter of design, and direct distribution is also included in the scope of the claims of the present invention.

In step S513, the information processing terminal 102 receives the proofread character string, and changes the "character string including misrecognition" already displayed on the display device of the information processing terminal 102 to the "proofread character string". replace.

Note that the display frames 404A to 404K in FIG. 4 may be edited separately for each utterance, or may be combined into one edit target. In order to prevent multiple proofreaders from proofreading one display frame 404 at the same time, the display frame 404 configured by one information processing terminal 102b may not be proofread by another information processing terminal 102b. . In addition, since the display frame 404 at the bottom of FIG. 4 continues to display a character string that is not segmented for speech recognition, it may not be possible to proofread. These are only design matters.

FIG. 7 is a diagram showing an example of a flow chart for explaining processing up to prioritization for analysis and proofreading of speech recognition results (step S508 in FIG. 5) according to the embodiment of the present invention. Each step of the flow chart of FIG. 7 is executed by the CPU 201 on the speech recognition server 101 .

In step S701, it is checked whether the recognition result of the voice data of the new utterance has been registered in the recognition result storage unit 520 or not. Specifically, assuming that up to 601J in FIG. 6 existed in the previous check, it is checked whether or not the next 601K is newly added. If registered ("Yes"), the process proceeds to step S702. If not registered ("No"), the process proceeds to step S704.

In step S702, morphological analysis is performed on the newly added character string of the speech recognition result. By the processing in step S702, the character string is classified and the part of speech is given as shown in the example of FIG. This generates a morpheme string. However, if the speech recognition result itself is given a part of speech by morphological analysis, step S702 is unnecessary and omitted.

In step S703, an individual name is extracted from the morpheme string. The technology for extracting the individual name is well-known technology such as Japanese Patent Laid-Open No. 2002-288190, so detailed description is omitted.

In step S704, among the recognition result character strings (eg, 601A to 601J in FIG. 6), the priority order to be corrected is set for those that have not yet been corrected. Details will be described later with reference to FIGS.

In step S705, if execution of the speech recognition system continues ("Yes"), the process returns to step S701. If execution of the speech recognition system has ended ("No"), the process of the flowchart of FIG. 7 is completed, and the process of the flowchart of FIG. 5 is returned to. In other words, the process returns to the state where step S508 in FIG. 5 has ended.

FIG. 8 is a diagram showing an example of a flowchart for explaining the prioritization process (step S704 in FIG. 7) according to the embodiment of the present invention. Each step of the flow chart of FIG. 8 is executed by the CPU 201 on the speech recognition server 101 .

Steps S801 to S808 are repeated processes for the results stored in the recognition result storage unit, that is, the results of speech recognition based on all speech data (for example, 10 speech data for 601A to 601J in FIG. 6).

In step S802, attention is paid to one speech recognition result. Specifically, focus on one of the 601A to 601J in order from the beginning.

In step S803, it is determined whether or not it is necessary to determine the priority of the speech recognition result of interest. It branches depending on whether or not it has already been proofread, or whether or not the condition described in the post-utterance condition 901 in FIG. 9 is satisfied. Since this determination is an OR condition of two types of determination, if either condition is satisfied, the result is "Yes", and the process proceeds to step S804. If none of the conditions are satisfied, the result is "No", and the process proceeds to step S805.

Of the above two conditions, whether or not proofreading has been completed will be specifically described in detail with reference to FIG. 10 (recognition statuses are given using the partial recognition result in FIG. 6 as an example). When the flow chart of FIG. 8 (that is, step S704 of FIG. 7) is executed for the first time after recognizing a certain segment of utterance, no decision has been made as to whether or not correction is necessary in FIG. The condition is not satisfied because there is no description (“No”). Regarding recognition results that have already been calibrated, the condition is satisfied (“Yes”) because “necessity of correction” in FIG. However, this part is a matter of design, and recognition results that have been calibrated once may also be subject to reprioritization. In that case, it is not set as "unnecessary" in S510.

A case in which the post-utterance condition 901 is used as the condition will be described. The intention of this condition is to judge that it is not useful to go back and proofread something that seems to have passed a certain period of time in chronological order after the completion of the utterance. . Three examples described in 901 in FIG. 9 will be specifically described.

The post-utterance condition 901 describes how to determine that a certain period of time has passed since the utterance. The conditions shown in FIG. 9 are merely examples, and any method for determining the passage of time other than these three methods is included in the present invention. Let's take an example one by one.

In example 1, in each utterance such as A to J in the example utterance 400 of FIG. 4, the actual time from the break where the utterance is considered to be completed is measured. In the example, if 180 seconds or more have passed since the end, the condition is that calibration is not required. The elapsed time is stored in the "elapsed time" field in the example of FIG.

Example 2 counts the number of characters, not the time. After the utterance is considered completed and separated, if 500 or more characters of the subsequent utterance are presented as recognition results, proofreading is not required at that point. and To explain using 603 in FIG. 6, if the total number of characters after 603B after 603A reaches 500 characters, the priority of the utterance of 601A is not calculated and proofreading is not required.

In Example 3, judgment is made based on how the reader sees it. The character string of the voice recognition result is normally not displayed on the reader's information processing terminal 102 as time passes. For example, the speech recognition result display screen 401 shown in FIGS. 4 and 11 is displayed in chronological order from the top at the breaks of utterances. ) can be conceived of as a user interface that scrolls and disappears upward (for example, the portion within the dotted line 1101 in FIG. 11).

Even if there is a different method and not in chronological order, i.e. if there is something that remains on the screen/disappears from the screen regardless of whether it is old or new, any reader can correct the errors of what has disappeared. is meaningless because it cannot be read. Therefore, it is conceivable that calibration will become unnecessary.

Although three examples are given here, other methods may be used. Also, these conditions may be combined (AND condition, OR condition).

In step S804, the "necessity of correction" of the information stored in FIG. 10 is set to "unnecessary" to make calibration unnecessary.

In step S805, the "correction necessity" in the proofreading status is set to "required" when proofreading is necessary, and the "correction status" is set to "not yet proofread".

Incidentally, as already explained, at the stage when proofreading is started in step S509 of FIG. , the correction status is changed to "completed" and the need for correction is changed to "unnecessary".

Next, in step S806, it is determined whether or not to recalculate the confidence factor, for example, according to 902 in FIG. Three conditions are described in 902 as examples, but the conditions are not limited to these.

For example, in example 1 of 902, it is determined whether or not the speech recognition result of interest includes a morpheme of a part of speech requiring confirmation or an individual name. For example, in B of FIG. 10, "number" is included, and this is registered as a confirmation required part of speech in 903 of FIG. In general, specific patterns containing numerals or numbers represent company sales, contracted amounts, dates, etc., so if there is an error, important information for the reader is not guaranteed. FIG. 10E also includes a numerical expression (1002) that is the result of individual name extraction. This is an individual name that needs to be checked for error even if it contains a specific person, organization, or numerical expression that is calibrated from a plurality of morphemes (904 in FIG. 9).

A second example of 902 is when the speech recognition result contains many morphemes with a particularly low degree of confidence. there is If the recognition confidence is low, there is a high possibility that many misrecognized morphemes are included, and thus the priority of proofreading is higher in a sense different from Example 1, in which there is individually important information.

Confidence is recalculated according to processing such as morphological analysis/individual name extraction and rules described in 902 and the like. The confidence recalculation method is described in the confidence recalculation method 905 as an example. That is, if important information is included in the above-described processing, the priority of proofreading is changed by changing the degree of certainty of the recognition result. For example, it describes how to recalculate the certainty when there are words registered in the part of speech to be confirmed 903 and information specified in the individual name extraction condition 904 (example 1 of 905, example 2).

Although the processing of morphological analysis in the flow chart of FIG. 8 is not described here, the speech recognition result itself is often divided into morpheme units, and the part of speech may be included as information of the speech recognition result. . If not included, morphological analysis or other methods (using a dictionary, etc.) may be used to assign parts of speech separately.

The same applies to individual name extraction. It may be included as a part of the embodiment of the present invention, or may include the result of individual name extraction on the voice recognition side as the voice recognition result.

In step S807, a calculation is performed to change the priority of the composition according to the time since the end of the speech. In the determination in step S803 and example 1 in step S901, it was decided that calibration was not required after a certain period of time had passed, but here the correspondence is to the recognition result before the certain period of time has passed. In other words, for example, if the recognition result is still displayed on the screen (example 901) without a certain period of time passing (example 1 of 901), it is close to the state where calibration is "unnecessary". The more time limit is left for proofreading, the higher the priority should be given to proofreading. The formula of example 3 of 905 lowers the confidence of the recognition result as time elapses.

In step S809, upon receiving the result of recalculation of the confidence as described above, the items for which proofreading is "required" are sorted according to the confidence, and information is provided to preferentially proofread items with a lower confidence. It is presented on the display device of the processing terminal 102 .

This completes the explanation of the processing according to the flowchart of FIG. Although the degree of certainty is changed according to certain rules here, it is not always necessary to change the degree of certainty. For example, another numerical value (such as a negative score) may be used that stores how much "scores" have been "deducted". The change in confidence is only an example and is a matter of design.

By performing the processing in FIG. 8, the priority order for proofreading is determined, and FIG. 10 shows an example in which the recognition result E has priority 1 and the recognition result I has priority 2, for example. . The proofreader determines the priority of proofreading based on this identifiable information. Alternatively, control may be performed so that only the items with the highest priority can be edited.

FIG. 11 is a diagram for showing an example of a user interface that displays speech recognition results according to the embodiment of the present invention. Although it is essentially the same diagram as FIG. 4, the following points are different.

1103e displays "medium" (under proofreading) indicating that one of the proofreaders is proofreading this recognition result. Priority orders 1 to 4 are assigned to 1103g to 1103j. This allows the proofreader to identify the order of priority for proofreading.

1103k is displayed as "present" because the speech recognition result of the current utterance is recognized halfway and the result is displayed. Even if this display frame can be calibrated, it may be controlled so that it cannot be calibrated until the next 1103l is displayed after the speech breaks.
This concludes the description of the present invention with reference to the drawings.

It goes without saying that the configuration and content of the various data described above are not limited to this, and may be configured in various configurations and content according to the application and purpose.

Although several embodiments have been described above, the present invention can be embodied as, for example, systems, devices, methods, computer programs or recording media. It may be applied to a system composed of, or may be applied to an apparatus composed of one device.

Further, the computer program in the present invention is a computer program capable of executing the processing methods of the flowcharts shown in FIGS. A computer program is stored that is computer executable for the method. The computer program in the present invention may be a computer program for each processing method of each device shown in FIGS.

As described above, a recording medium recording a computer program that realizes the functions of the above-described embodiments is supplied to a system or apparatus, and a computer (or CPU or MPU) of the system or apparatus is stored in the recording medium. Needless to say, the object of the present invention can also be achieved by reading and executing the program.

In this case, the computer program itself read from the recording medium implements the novel functions of the present invention, and the recording medium storing the computer program constitutes the present invention.

Examples of recording media for supplying computer programs include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, CD-Rs, DVD-ROMs, magnetic tapes, non-volatile memory cards, ROMs, EEPROMs, A silicon disk, a solid state drive, or the like can be used.

In addition, by executing a computer program read by a computer, not only the functions of the above-described embodiments are realized, but also the OS (operating system) etc. running on the computer based on the instructions of the computer program. Needless to say, a case where part or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing are included.

Furthermore, after the computer program read from the recording medium is written in the memory provided in the function expansion board inserted into the computer or the function expansion unit connected to the computer, the function is executed based on the instructions of the computer program code. Needless to say, a case where a CPU or the like provided in an expansion board or function expansion unit performs part or all of the actual processing and the processing implements the functions of the above-described embodiments.

Moreover, the present invention may be applied to a system composed of a plurality of devices or to an apparatus composed of a single device. Moreover, it goes without saying that the present invention can be applied to a case where it is achieved by supplying a computer program to a system or apparatus. In this case, by loading a recording medium storing a computer program for achieving the present invention into the system or apparatus, the system or apparatus can enjoy the effects of the present invention.

Furthermore, by downloading and reading out the computer program for achieving the present invention from a server, database, etc. on the network using a communication program, the system or apparatus can enjoy the effects of the present invention.
It should be noted that all configurations obtained by combining each of the above-described embodiments and modifications thereof are also included in the present invention.

101 speech recognition server 102 information processing terminal 320 recognition result storage unit 321 speech data reception unit 322 speech recognition unit 323 recognition result transmission unit 324 recognition result management unit 325 priority determination unit 326 proofreading result reception unit 327 proofreading result distribution unit

Claims (9)

An information processing device for acquiring a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data,
Output for causing a display device to display information indicating a priority order in which the series of text data for each of the segmented voice data should be corrected based on the first certainty of the series of text data. a control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The output control means indicates the order of priority for proofreading the series of text data based on the first certainty adjusted based on whether the character data included in the series of text data is a predetermined part of speech. An information processing apparatus characterized by displaying information on a display device . An information processing device for acquiring a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data,
Output for causing a display device to display information indicating a priority order in which the series of text data for each of the segmented voice data should be corrected based on the first certainty of the series of text data. a control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The output control means sets a priority order for proofreading the series of text data based on the first degree of certainty adjusted based on a determination as to whether character data included in the series of text data indicates an individual name. 1. An information processing apparatus, characterized in that the information to be displayed is displayed on a display device . An information processing device for acquiring a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data,
Output for causing a display device to display information indicating a priority order in which the series of text data for each of the segmented voice data should be corrected based on the first certainty of the series of text data. a control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The information processing apparatus, wherein the output control means causes the display device to display information to the effect that correction of the series of text data is unnecessary based on the time elapsed since the utterance. 4. The series of text data according to any one of claims 1 to 3 , characterized in that the series of text data is text data including at least one character data obtained by recognizing speech data divided by utterance breaks. information processing equipment. An information processing device and a display device for obtaining text data based on recognition of a series of speech data segmented from continuous speech data according to a predetermined condition, and a first certainty factor indicating the likelihood of recognition of the series of text data. An information processing system comprising
causing the display device to display information indicating a priority order for correcting the series of text data based on a first certainty factor of the series of text data, with respect to the series of text data for each of the divided voice data; an output control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The output control means indicates the order of priority for proofreading the series of text data based on the first certainty adjusted based on whether the character data included in the series of text data is a predetermined part of speech. An information processing system characterized by displaying information on a display device . An information processing device and a display device for obtaining text data based on recognition of a series of speech data segmented from continuous speech data according to a predetermined condition, and a first certainty factor indicating the likelihood of recognition of the series of text data. An information processing system comprising
causing the display device to display information indicating a priority order for correcting the series of text data based on a first certainty factor of the series of text data, with respect to the series of text data for each of the divided voice data; an output control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The output control means sets a priority order for proofreading the series of text data based on the first degree of certainty adjusted based on a determination as to whether character data included in the series of text data indicates an individual name. 1. An information processing system characterized by displaying information to be displayed on a display device . An information processing device and a display device for obtaining text data based on recognition of a series of speech data segmented from continuous speech data according to a predetermined condition, and a first certainty factor indicating the likelihood of recognition of the series of text data. An information processing system comprising
causing the display device to display information indicating a priority order for correcting the series of text data based on a first certainty factor of the series of text data, with respect to the series of text data for each of the divided voice data; an output control means;
Receiving means for receiving correction of the text data in order to update the text data in which the information is displayed ,
The information processing system, wherein the output control means causes the display device to display information to the effect that correction of the series of text data is unnecessary based on the time elapsed since the utterance. A control method for an information processing device for acquiring a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data. There is
The output control means, for the series of text data for each of the divided voice data,
an output control step of causing a display device to display information indicating a priority order for correcting the series of text data based on the first certainty factor of the series of text data;
a receiving step in which the receiving means receives a correction of the text data in order to update the text data in which the information is displayed ;
In the output control step, a priority order for proofreading the series of text data is indicated based on the first certainty adjusted based on whether the character data included in the series of text data is a predetermined part of speech. A control method for an information processing device, comprising : displaying information on a display device . Executable in an information processing device for obtaining a series of text data based on recognition of speech data segmented from continuous speech data according to a predetermined condition and a first certainty factor indicating the likelihood of recognition of the series of text data a program,
the information processing device,
Output for causing a display device to display information indicating a priority order in which the series of text data for each of the segmented voice data should be corrected based on the first certainty of the series of text data. a control means;
In order to update the text data in which the information is displayed, it functions as a receiving means for receiving correction of the text data,
The output control means indicates the order of priority for proofreading the series of text data based on the first certainty adjusted based on whether the character data included in the series of text data is a predetermined part of speech. A program characterized by displaying information on a display device .

Images