Information processing apparatus and maas providing method

Abstract

To improve a response speed in an interactive system.SOLUTION: The information processing device 10 is an information processing device including a controller 11 that acquires utterance information indicating an utterance of an utterer in real time, performs prediction conversion on the utterance to generate a plurality of predicted utterance candidates, generates a plurality of response sentence candidates for each of the plurality of predicted utterance candidates, specifies, when the utterance is completed, a predicted utterance candidate similar to the completed utterance among the plurality of predicted utterance candidates as a target predicted utterance, and selects a response sentence candidate corresponding to the specified target predicted utterance among the plurality of response sentence candidates as a target response sentence. The selected target response sentence is output to the speaker.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device and a MaaS providing method. [Background technology]

[0002] Dialogue systems have been known for some time. For example, Patent Document 1 discloses a dialogue device that generates multiple response candidates in parallel in response to an input from a user and responds with the one with the highest evaluation value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-165746 Summary of the Invention [Problem to be solved by the invention]

[0004] There is a demand for further improvement in response speed in dialogue systems. [Means for solving the problem]

[0005] An information processing device according to the present disclosure is an information processing device including a control unit, which acquires utterance information indicating a speaker's utterance in real time, performs predictive conversion on the utterance to generate a plurality of predicted utterance candidates, generates a plurality of response sentence candidates for each of the plurality of predicted utterance candidates, and when the utterance is completed, identifies a predicted utterance candidate from among the plurality of predicted utterance candidates that is closest to the completed utterance as a target predicted utterance, selects a response sentence candidate from among the plurality of response sentence candidates that corresponds to the identified target predicted utterance as a target response sentence, and outputs the selected target response sentence to the speaker. [Effects of the Invention]

[0006] According to the present disclosure, it is possible to improve the response speed in a dialogue system. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a diagram showing a schematic configuration of a system according to an embodiment of the present invention. [Figure 2] 4 is a flowchart showing the operation of the system according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings.

[0009] The configuration of an information processing device 10 according to this embodiment will be described with reference to Fig. 1. The information processing device 10 is, for example, a server device, a computer such as a PC (Personal Computer), or a general-purpose or dedicated electronic device. The information processing device 10 may also be a server belonging to a cloud computing system.

[0010] First, an overview of this embodiment will be described, and details will be described later. The information processing device 10 acquires utterance information indicating an utterance of a speaker in real time, performs predictive conversion on the utterance, and generates a plurality of predicted utterance candidates. The information processing device 10 generates a plurality of response sentence candidates for each of the plurality of predicted utterance candidates, and when the utterance is completed, identifies a predicted utterance candidate that is closest to the completed utterance from the plurality of predicted utterance candidates as a target predicted utterance. The information processing device 10 selects a response sentence candidate corresponding to the identified target predicted utterance from the plurality of response sentence candidates as a target response sentence, and outputs the selected target response sentence to the speaker.

[0011] According to this embodiment, it is possible to create a plurality of predicted utterance candidates in advance, predict what the speaker will say, and then generate a plurality of response sentence candidates. Therefore, it is possible to output a response sentence more quickly than when a response sentence is generated after all utterances have been completed. Therefore, it is possible to improve the response speed in the dialogue system.

[0012] The configuration of the information processing device 10 will be described with reference to Fig. 1. The information processing device 10 includes a control unit 11, a storage unit 12, a communication unit 13, an input unit 14, and an output unit 15.

[0013] The control unit 11 includes at least one processor, at least one programmable circuit such as an FPGA (field-programmable gate array), at least one dedicated circuit such as an ASIC (application specific integrated circuit), or any combination thereof. The processor is a general-purpose processor such as a CPU (central processing unit) or a GPU (graphics processing unit), or a dedicated processor specialized for a specific process. The control unit 11 controls each unit of the information processing device 10 and executes processes related to the operation of the information processing device 10.

[0014] The storage unit 12 includes at least one semiconductor memory, at least one magnetic memory, at least one optical memory, or any combination thereof. The semiconductor memory is, for example, a random access memory (RAM), a read only memory (ROM), or a flash memory. The RAM is, for example, a static random access memory (SRAM) or a dynamic random access memory (DRAM). The ROM is, for example, an electrically erasable programmable read only memory (EEPROM). The flash memory is, for example, a solid-state drive (SSD). The magnetic memory is, for example, a hard disk drive (HDD). The storage unit 12 functions as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 12 stores information used in the operation of the information processing device 10 and information obtained by the operation of the information processing device 10.

[0015] The communication unit 13 includes at least one communication module. The communication module is, for example, a module compatible with a wired LAN (local area network) communication standard such as Ethernet (registered trademark), a wireless LAN communication standard such as IEEE (Institute of Electrical and Electronics Engineers) 802.11, or a mobile communication standard such as LTE (Long Term Evolution), 4G (4th generation) standard, or 5G (5th generation) standard. The communication unit 13 receives information used in the operation of the information processing device 10 and transmits information obtained by the operation of the information processing device 10. The communication unit 13 enables the information processing device 10 to transmit and receive information to and from other devices via a network.

[0016] The network may include the Internet, at least one wide area network (WAN), at least one metropolitan area network (MAN), or a combination thereof. The network may include at least one wireless network, at least one optical network, or a combination thereof. The wireless network may be, for example, an ad-hoc network, a cellular network, a wireless LAN, or a satellite communication network.

[0017] The input unit 14 is, for example, a microphone, physical keys, capacitance keys, a pointing device, or a touch screen integrated with a display. The input unit 14 accepts an operation to input data used for the operation of the information processing device 10. The input unit 14 may be connected to the information processing device 10 as an external input device instead of being provided in the information processing device 10. As a connection interface, an interface compatible with standards such as USB (Universal Serial Bus) or Bluetooth (registered trademark) may be used.

[0018] The output unit 15 is, for example, a speaker or a display. The display is, for example, an LCD (liquid crystal display) or an organic EL (electro luminescent) display. The output unit 15 outputs data obtained by the operation of the information processing device 10. The output unit 15 may be connected to the information processing device 10 as an external output device instead of being provided in the information processing device 10. As a connection interface, an interface compatible with standards such as USB or Bluetooth (registered trademark) may be used.

[0019] The functions of the information processing device 10 are realized by executing a program according to this embodiment on a processor serving as the control unit 11. That is, the functions of the information processing device 10 are realized by software. The program causes a computer to execute the operations of the information processing device 10, thereby causing the computer to function as the information processing device 10. That is, the computer functions as the information processing device 10 by executing the operations of the information processing device 10 in accordance with the program.

[0020] The program may be stored in a non-transitory computer-readable medium, such as a flash memory, a magnetic recording device, an optical disk, a magneto-optical recording medium, or a ROM.

[0021] Some or all of the functions of the information processing device 10 may be realized by a programmable circuit or a dedicated circuit as the control unit 11. In other words, some or all of the functions of the information processing device 10 may be realized by hardware.

[0022] The operation of the information processing device 10 according to this embodiment will be described with reference to Fig. 2. In the following, communication between the information processing device 10 and an external device is performed via the communication unit 13 and a network.

[0023] In S1 of FIG. 2, the control unit 11 acquires speech information indicating the speaker's speech in real time. Any method may be adopted to acquire the speech information. For example, the control unit 11 accepts audio input of the speaker's speech via a microphone serving as the input unit 14. The control unit 11 converts the speech into a character string and acquires the speech information representing the speech as a character string. However, the present invention is not limited to this. The control unit 11 may communicate with a terminal device used by the speaker, receive information including the speech input by the speaker to the terminal device from the terminal device, and acquire the speech information based on the speech. Any audio processing technology may be adopted to convert into a character string.

[0024] In S2, the control unit 11 performs predictive conversion on the utterance based on the utterance information to generate multiple predicted utterance candidates. Any natural language processing technology may be used for the predictive conversion. The control unit 11 sequentially generates predicted utterance candidates in real time for the content of the utterance up to the point when the utterance information was acquired, even while the speaker is speaking. In this example, the utterance indicated by the utterance information is the character string "a delicious restaurant," and the control unit 11 generates multiple character strings such as "I found a delicious restaurant" and "Where is a delicious restaurant?" as predicted utterance candidates.

[0025] In S3, the control unit 11 generates a plurality of response sentence candidates for each of the plurality of predicted utterance candidates. Any natural language processing technology may be used to generate the response sentence candidates. In this example, the control unit 11 generates a response sentence candidate "Where is it? Please tell me," in response to the predicted utterance candidate "I found a delicious restaurant," and generates a response sentence candidate "I recommend Store A in front of the station," in response to the predicted utterance candidate "Where is a delicious restaurant?"

[0026] In S4, the control unit 11 acquires an announcement to be notified to the speaker and assigns a priority to the announcement. The announcement includes a message to alert the speaker.

[0027] Any method may be used to acquire the announcement text. For example, the control unit 11 communicates with a camera installed near a microphone serving as the input unit 14 through which the speaker inputs voice, and acquires an image of the speaker captured by the camera. The control unit 11 acquires the announcement text when it detects a change in the situation around the speaker from the image. Any image analysis method may be used to detect the change in the situation around the speaker. Examples of changes in the situation around the speaker include the opening and closing of a door of a building or the like near the speaker, or the passing of another pedestrian, such as a wheelchair user, near the speaker. The control unit 11 may read and acquire the announcement text corresponding to the detected change from the memory unit 12. The control unit 11 may also acquire the announcement text by receiving it from an external server device. For example, the control unit 11 may communicate with a terminal device used by an operator who remotely monitors the speaker by viewing camera footage, and receive the announcement text input by the operator from the terminal device.

[0028] The control unit 11 assigns a priority to the acquired announcement. Any method may be used to assign a priority. For example, the control unit 11 refers to a priority dictionary in which a priority for each word is preset, determines a priority to be assigned based on the priority dictionary, and assigns information indicating the priority to the announcement. The priority dictionary may be stored in the storage unit 12. For example, the control unit 11 may identify a word with the highest priority among one or more words contained in the announcement, and determine the priority of the identified word as the priority of the announcement. If a higher value is set in the priority dictionary as the priority becomes higher, for example, the control unit 11 may determine the priority of the announcement to be a priority calculated by adding up the priorities of one or more words contained in the announcement. The control unit 11 may assign a uniform, preset priority to all announcements.

[0029] In this example, the control unit 11 detects a change in the image of the speaker, namely, a door opening near the speaker, and acquires an announcement saying, "Please move away from the door." The control unit 11 assigns a priority of 10 to the announcement.

[0030] In S5, when the speaker completes his / her utterance, the control unit 11 identifies, from among the plurality of predicted utterance candidates, a predicted utterance candidate that is closest to the completed utterance as a target predicted utterance. The control unit 11 may determine that the speaker's utterance has been completed when a predetermined time has elapsed since the speaker stopped speaking. For example, the control unit 11 may determine the degree of approximation to the completed utterance using any language processing technology and identify the target predicted utterance. In this example, it is assumed that the completed utterance indicated by the utterance information is "Are there any good restaurants?". The control unit 11 identifies, from among the plurality of predicted utterance candidates generated in S2, "Where are there any good restaurants?" as the target predicted utterance.

[0031] In S6, the control unit 11 selects, as a target response sentence, a response sentence candidate corresponding to the identified target predicted utterance from among the plurality of response sentence candidates. In this example, the control unit 11 selects, as a target response sentence, the response sentence candidate "I recommend Store A in front of the station" corresponding to the target predicted utterance "Where is a good restaurant?"

[0032] In S7, the control unit 11 assigns a priority to the target response sentence. Any method may be used to assign a priority. For example, the control unit 11 determines the priority to be assigned based on a priority dictionary and assigns information indicating the priority to the target response sentence. In this example, the control unit 11 assigns a priority of 5 to the target response sentence "I recommend Store A in front of the station."

[0033] In S8, the control unit 11 compares the priority of the target response sentence with the priority of the announcement sentence, and outputs the one with the higher priority via the output unit 15. If the priority of the announcement sentence is higher than the priority of the target response sentence, the control unit 11 outputs the announcement sentence with priority over the target response sentence. If the priority of the announcement sentence is equal to or lower than the priority of the target response sentence, the control unit 11 outputs the response sentence with priority over the announcement sentence. The control unit 11 may output audio via a speaker serving as the output unit 15. The control unit 11 may output the target response sentence or the announcement sentence by communicating with the speaker's terminal device and transmitting audio information indicating the target response sentence or the announcement sentence to the terminal device. Thereafter, the operation of the information processing device 10 ends.

[0034] In this example, the higher the priority, the larger the value assigned. The priority assigned to the announcement sentence "Please move away from the door" is 10, and the priority assigned to the target response sentence "Where is a good restaurant?" is 5. The control unit 11 outputs the announcement sentence to the speaker, prioritizing it over the target response sentence. This makes it possible to flexibly notify the speaker of announcement sentences, such as warnings, even during a conversation with the speaker.

[0035] In one variation of the present disclosure, when the control unit 11 acquires instruction information indicating a user's instruction to output an announcement, the control unit 11 may output the announcement acquired in S4 above over the target response sentence, regardless of the above-mentioned priority. The user may, for example, be an operator who remotely monitors the speaker by watching video from a camera capturing the speaker. The control unit 11 may communicate with a terminal device used by the user and receive instruction information from the terminal device. Even if the control unit 11 is in the middle of outputting the target response sentence in S8, when the control unit 11 acquires the instruction information, the control unit 11 may interrupt the output of the target response sentence and output the announcement. According to this variation, the announcement can be output at an appropriate timing based on manual judgment.

[0036] In one variation of the present disclosure, the control unit 11 may acquire an image in which the speaker's belongings are reflected and generate predicted utterance candidates based on the acquired image. In this variation, the control unit 11 acquires an image in which the speaker's belongings are reflected, captured by a camera installed near a microphone serving as the input unit 14. The control unit 11 may communicate with a terminal device to which the speaker inputs voice and receive and acquire images captured by the camera of the terminal device. The control unit 11 extracts the belongings from the acquired image using any image analysis processing technology. The control unit 11 generates multiple predicted utterance candidates, including predicted utterance candidates containing character strings indicating the belongings. The belongings may include the speaker's belongings, such as a suitcase, or children the speaker is carrying. For example, the control unit 11 acquires an image captured by a camera capturing the speaker and detects "children" as the speaker's belongings using any image analysis technology. In this case, the control unit 11 may generate predicted utterance candidates such as "Where are some good restaurants I can go to with my children?" According to this modification, it is possible to generate predicted utterance candidates that are more likely to resemble the utterance of the speaker.

[0037] The control unit 11 may assign a priority to each response sentence candidate each time the control unit 11 generates the response sentence candidates in S3 above. The control unit 11 may use the assigned priority as the priority of the target response sentence in S7 above.

[0038] The present disclosure is not limited to the above-described embodiments. For example, two or more blocks shown in the block diagram may be integrated, or one block may be divided. Instead of executing two or more steps shown in the flowchart in chronological order as described, the steps may be executed in parallel or in a different order depending on the processing capabilities of the device executing each step or as needed. Other modifications are possible without departing from the spirit of the present disclosure. Furthermore, for example, the information processing device 10 according to the above-described embodiment may be used to provide MaaS (Mobility as a Service), a service utilizing mobility. [Explanation of symbols]

[0039] 10. Information processing equipment 11 Control section 12 Storage section 13 Communications Department 14 Input section 15 Output section

Claims

1. An information processing device including a control unit, The control unit Acquires speech information indicating the speaker's speech in real time, performing predictive conversion on the utterance to generate a plurality of predicted utterance candidates; generating a plurality of response sentence candidates for each of the plurality of predicted utterance candidates; When the utterance is completed, a predicted utterance candidate that is closest to the completed utterance is identified as a target predicted utterance from among the plurality of predicted utterance candidates; selecting, from the plurality of response sentence candidates, a response sentence candidate corresponding to the identified target predicted utterance as a target response sentence; An information processing device that outputs the selected target response sentence to the speaker.

2. The control unit assigning a priority to the target response sentence; Acquire an announcement to be notified to the speaker, and assign a priority to the announcement; comparing the priority of the target response sentence with the priority of the announcement sentence; The information processing device according to claim 1 , wherein when the priority of the announcement sentence is higher than the priority of the target response sentence, the announcement sentence is output with priority over the target response sentence.

3. The information processing device according to claim 2 , wherein the control unit acquires an image in which the speaker is reflected, and acquires the announcement text when a change in a situation around the speaker is detected from the image.

4. 4. The information processing device according to claim 2, wherein when the control unit acquires instruction information indicating a user instruction to output the announcement sentence, the control unit outputs the acquired announcement sentence in priority to the target response sentence regardless of the priority.

5. 4. The information processing device according to claim 1, wherein the control unit acquires an image in which an accessory of the speaker is reflected, and generates the plurality of predicted utterance candidates based on the image, the predicted utterance candidates including a character string indicating the accessory.

6. A method for providing MaaS (Mobility as a Service) using the information processing device according to claim 1.

Images