Information processing device, information processing method, program, recording medium, and data structure

Abstract

To provide an information processing apparatus configured to easily retrieve information on a target that draws the interest during travel, an information processing method, a program, a recording medium, and a data structure.SOLUTION: An information processing apparatus includes: a voice acquisition unit which acquires input voice input through a speech of a user; a speech position acquisition unit which acquires a speech position at which the user speaks; a map data acquisition unit which acquires map data; a specifying unit which specifies one or more names of places or features based on the speech position and the map data; and a search word generation unit which generates a search word including a first word to be generated based on the input voice and a second word to be generated based on the one or more names of places or features specified by the specifying unit.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, a program, a recording medium, and a data structure.

Background Art

[0002] There is known a navigation apparatus that recognizes voice input from a user during movement in a vehicle or the like and gives an answer to the recognized content. For example, in Patent Document 1, when there is a voice input from a user such as "What is the building on the left?", based on the result of voice recognition, the position of the own vehicle, the traveling direction of the own vehicle, and the map database, facilities existing on the left of the own vehicle are searched, and a navigation apparatus that guides the name of the facility by voice is disclosed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the navigation apparatus described in Patent Document 1, for example, when a plurality of buildings existing on the left of the own vehicle are hit as the above search result, all the hit buildings are guided to the user as candidates. Therefore, one of the problems is that the user is guided to buildings that have little relevance to the building that the user has questioned, and the information that the user wants to know is buried in the search results.

[0005] In addition, in the navigation apparatus described in Patent Document 1, when a building that the user is interested in is not registered in the above map database, there is a problem that information about the object cannot be obtained.

[0006] The present invention has been made in view of the above-mentioned points, and aims to provide an information processing device, an information processing method, a program, a recording medium, and a data structure that enable easy retrieval of information about an object of interest while on the move. [Means for solving the problem]

[0007] The information processing device according to claim 1 is characterized by comprising: a voice acquisition unit that acquires input voice input by a user's utterance; a voice position acquisition unit that acquires the voice position at which the utterance was made by the user; a map data acquisition unit that acquires map data; a specification unit that identifies one or more place names or features based on the voice position and the map data; and a search word generation unit that generates a search word including a first word generated based on the input voice and a second word generated based on the one or more place names or features identified by the specification unit.

[0008] The information processing method described in claim 9 is an information processing method performed by an information processing device, comprising: a speech acquisition step of acquiring input speech input by a user's utterance; a speech position acquisition step of acquiring the utterance position where the utterance was made by the user; a map data acquisition step of acquiring map data; a specification step of identifying one or more place names or features based on the utterance position and the map data; and a search word generation step of generating a search word that includes a first word generated based on the input speech and a second word generated based on the one or more place names or features identified by the specification unit.

[0009] The program described in claim 10 is a program to be executed by a computer, characterized in that it causes the computer to execute: a voice acquisition step of acquiring input voice input by a user's utterance; an utterance position acquisition step of acquiring the utterance position where the utterance was made by the user; a map data acquisition step of acquiring map data; an identification step of identifying one or more place names or features based on the utterance position and the map data; and a search word generation step of generating a search word that includes a first word generated based on the input voice and a second word generated based on the one or more place names or features identified by the identification unit.

[0010] The recording medium described in claim 11 is a recording medium on which the above program is recorded.

[0011] The data structure described in claim 12 is a data structure used in an information processing device having a specification unit and a search word generation unit, the data structure being used in a process in which the specification unit specifies one or more place names or features based on the utterance of a user and the utterance location where the utterance was made by the user, the specification unit specifies one or more place names or features based on the utterance location and map data, and the search word generation unit generates a search word that includes a first word generated based on the input voice and a second word generated based on the one or more place names or features. [Brief explanation of the drawing]

[0012] [Figure 1] This figure shows the configuration of the front seat area of ​​the vehicle according to Embodiment 1 and the view from the front of the vehicle. [Figure 2] This is a block diagram showing an example of the configuration of the information processing device according to Example 1. [Figure 3] This figure shows the configuration of search words associated with the utterance location on the map according to Example 1. [Figure 4] This is a flowchart showing the control routine of the information processing device according to Example 1. [Figure 5] It is a flowchart showing a control subroutine of the information processing apparatus according to Embodiment 1. [Figure 6] It is a diagram showing the configuration of the front seat portion of the vehicle and the forward view of the vehicle according to Embodiment 2. [Figure 7] It is a block diagram showing an example of the configuration of the information processing apparatus according to Embodiment 2. [Figure 8] It is a diagram showing an example of a method for specifying a place name by the information processing apparatus according to Embodiment 2. [Figure 9] It is a table showing examples of nouns classified by genre according to Embodiment 2 and the corresponding distance ranges. [Figure 10] It is a table showing examples of expressions of directions according to Embodiment 2 and the corresponding angular ranges. [Figure 11] It is a flowchart showing a control subroutine of the information processing apparatus according to Embodiment 2. [Figure 12] It is a diagram showing the configuration of the front seat portion of the vehicle and the forward view of the vehicle according to Embodiment 3. [Figure 13] It is a diagram showing an example of a method for specifying a place name by the information processing apparatus according to Embodiment 3. [Figure 14] It is a table showing examples of expressions of directions according to Embodiment 3 and the corresponding angular ranges. [Figure 15] It is a flowchart showing a control subroutine of the information processing apparatus according to Embodiment 3. [Figure 16] It is a diagram showing an example of a method for specifying a ground object by the information processing apparatus according to Embodiment 4. [Figure 17] It is a flowchart showing a control subroutine of the information processing apparatus according to Embodiment 4. [Figure 18] It is a diagram showing an example of a method for specifying a ground object by the information processing apparatus according to Embodiment 5. [Figure 19] It is a flowchart showing a control subroutine of the information processing apparatus according to Embodiment 5.

Mode for Carrying Out the Invention

[0013] Hereinafter, embodiments of the present invention will be specifically described with reference to the drawings. In the drawings, the same components are denoted by the same reference numerals, and the description of overlapping components will be omitted.

Embodiment

[0014] FIG. 1 is a diagram showing the front seat portion of a vehicle M as a moving body according to Embodiment 1 and the front view of the vehicle M. The in-vehicle device 10 is connected to each of a microphone 12, a GPS receiver 13, a touch panel display 14, and a speaker 15 installed in the vehicle M, and includes a control unit that controls these. The in-vehicle device 10 is disposed, for example, at the center in the dashboard DB of the front seat of the vehicle M.

[0015] The microphone 12 is a voice input device that receives sounds inside the vehicle, for example, the voice uttered by the driver of the vehicle M, and converts it into an electrical signal. In the present embodiment, the microphone 12 is disposed on the dashboard DB.

[0016] The GPS receiver 13 is a receiver that receives signals (GPS signals) from GPS (Global Positioning System) satellites. The GPS receiver 13 transmits the received GPS signal to the in-vehicle device 10. In the present embodiment, the GPS receiver 13 is disposed on the dashboard DB.

[0017] The touch panel display 14 is a display device in which a display that performs screen display based on the control of the in-vehicle device 10 and a touch panel that receives input operations from the passengers of the vehicle M are combined. For example, a navigation image in which the current position of the vehicle M is superimposed on a map is displayed on the touch panel display 14. In the present embodiment, the touch panel display 14 is disposed at the center of the dashboard DB.

[0018] The speaker 15 is a voice output device that outputs voice based on the control of the in-vehicle device 10. In the present embodiment, the speakers 15 are provided on each of the two A-pillars AP.

[0019] The positions of the in-vehicle device 10, microphone 12, GPS receiver 13, touch panel display 14, and speaker 15 in the front seat area of ​​vehicle M described above are merely examples, and they may be placed in other locations.

[0020] Figure 1 shows a ramen shop with a queue of people visible from inside vehicle M, to the right and slightly ahead of the vehicle's direction of travel. In Figure 1, the passenger in vehicle M (for example, the driver) is curious about what kind of ramen shop it is and asks the question "What kind of ramen shop is that with the queue?" into microphone 12.

[0021] In this embodiment, the in-vehicle device 10 is configured to recognize voice questions from the user of the in-vehicle device 10, for example, the driver of vehicle M. That is, the in-vehicle device 10 converts the voice input from the driver of vehicle M into text consisting of strings. The in-vehicle device 10 extracts a predetermined type of text from the converted text and generates a first word as a string using the extracted text.

[0022] Furthermore, in this embodiment, the in-vehicle device 10 is configured to acquire the position of vehicle M, i.e., the utterance position, when voice input is received through the utterance of a passenger in vehicle M. Based on the utterance position and map data, the in-vehicle device 10 identifies the place name of the utterance position on the map and generates a second word as a string using the identified place name.

[0023] In this embodiment, the in-vehicle device 10 combines the first word and the second word described above to generate a search term for searching for a target TA that the user is interested in, and registers the search term in association with the utterance location on the map.

[0024] Figure 2 is a block diagram showing an example of the configuration of the in-vehicle device 10 according to Embodiment 1. In this embodiment, the in-vehicle device 10 is a device in which a control unit 17, a communication unit 18, an input unit 19, an output unit 21, and a large-capacity storage device 22 cooperate via a system bus 16.

[0025] The control unit 17 is composed of a CPU (Central Processing Unit) 17A, ROM (Read Only Memory) 17B, RAM (Random Access Memory) 17C, etc., and functions as a computer. The CPU 17A reads and executes various programs stored in the ROM 17B and the mass storage device 22 to realize various functions.

[0026] Various programs may be acquired, for example, from other server devices via a network, or they may be recorded on a recording medium and read via various drive devices. In other words, various programs stored in the mass storage device 22 (including programs for executing processing in the in-vehicle device 10 described later) can be transmitted via a network, and can also be recorded on a computer-readable recording medium and transferred.

[0027] The communication unit 18 is a communication device that transmits and receives data with external devices according to instructions from the control unit 17. The communication unit 18 is, for example, a NIC (Network Interface Card) for connecting to a network.

[0028] In this embodiment, the control unit 17 is connected to a server computer that functions as a search engine such as Google® or Yahoo!® via the communication unit 18. The in-vehicle device 10 and the server computer can send and receive data to and from each other using a communication protocol such as TCP / IP. The connection between the in-vehicle device 10 and the network can be made by mobile communication such as 3G (3rd Generation), 4G (4th Generation), or 5G (5th Generation), or by wireless communication such as Wi-Fi®.

[0029] The input unit 19 is an interface unit that enables communication between the control unit 17 of the in-vehicle device 10, the microphone 12, the GPS receiver 13, and the touch panel display 14. In this embodiment, the control unit 17 functions as an audio acquisition unit that acquires the voice of the vehicle M driver input from the microphone 12 via the input unit 19.

[0030] Furthermore, in this embodiment, the control unit 17 functions as a speech position acquisition unit that acquires the position of the vehicle M, i.e., the speech position described above, based on the GPS signal from the GPS receiver 13 when voice input is received from the passenger of the vehicle M via the input unit 19.

[0031] The output unit 21 is an interface unit that is communicatively connected to the touch panel display 14 and the speaker 15. The control unit 17 can transmit video or image signals to the touch panel display 14 via the output unit 21 to display them, or transmit audio signals to the speaker 15 to output sound.

[0032] The large-capacity storage device 22 is composed of, for example, a hard disk drive, an SSD (solid state drive), flash memory, etc., and is a storage device that stores various programs such as operating systems and terminal software.

[0033] The speech recognition database (hereinafter referred to as the speech recognition DB) 22A of the large-capacity storage device 22 is a database that stores language models, pronunciation dictionaries, etc., for use in speech recognition of input speech. When speech is input from the user, the control unit 17 refers to the speech recognition DB 22A to perform speech recognition and converts the speech into text consisting of strings.

[0034] The map information database (hereinafter referred to as the map information DB) 22B of the large-capacity storage device 22 is a database in which map information is stored. The control unit 17 identifies place names and features on the map by referring to the above-mentioned speech location and the map information stored in the map information DB 22B. In other words, in this embodiment, the control unit 17 is a map data acquisition unit that acquires map information (map data).

[0035] The following describes an example of the search word generation process performed by the control unit 17 in this embodiment.

[0036] [Extracting text] First, let's explain the text extraction. In this embodiment, the control unit 17 converts the voice input from the passenger of vehicle M into text consisting of strings using the voice recognition DB22A described above. Specifically, the control unit 17 performs voice recognition on the voice input from the driver of vehicle M, "What kind of ramen shop is that with the long line?", and converts it into text.

[0037] The control unit 17 then extracts common nouns and words indicating attributive modifiers, attributive phrases, or attributive clauses related to those common nouns from the converted text. Specifically, from the string "What kind of shop is that ramen shop with a long line?" which has been converted into text by speech recognition, the control unit 17 extracts "ramen shop," which is a common noun representing a general name, and "that long line," which is a word indicating an attributive clause related to "ramen shop."

[0038] [Generating the first word] Next, the generation of the first word will be explained. In this embodiment, the control unit 17 defines the extracted common noun and the word indicating the attributive clause related to the common noun as the first word.

[0039] In this embodiment, the control unit 17 excludes so-called demonstrative pronouns such as "this, that, over there, which" and "this, that, over there, which" from the first word when a common noun or a word indicating an attributive modifier, attributive phrase, or attributive clause related to that common noun is included. In other words, in this embodiment, the control unit 17 defines "ramen shop" and "there is a line" as the first word.

[0040] [Generating the second word] Next, the generation of the second word will be described. In this embodiment, the control unit 17 identifies the place name of the utterance location of the passenger in the vehicle M based on the utterance location and the map information stored in the map information DB 22B. In other words, the control unit 17 functions as a identifying unit that identifies a place name based on the utterance location of the passenger in the vehicle M and the map data.

[0041] In this embodiment, the control unit 17 identifies the address of the utterance location on the map. Specifically, the control unit 17 identifies the address of the utterance location on the map, for example, "Shinjuku-ku △△-cho ○-chome ×-ban □-go". In this embodiment, the control unit 17 uses the word indicating the town name of the identified utterance location address as the second word. That is, in this embodiment, the control unit 17 uses "Shinjuku-ku △△-cho" as the second word.

[0042] [Generating search terms] Next, the generation of search words will be explained. The control unit 17 generates search words by combining the first word and the second word that have been generated. In other words, the control unit 17 is a search word generation unit that generates search words using the first word and the second word.

[0043] Specifically, the control unit 17 combines the first words, "a line has formed" and "ramen shop," with the second word, "Shinjuku-ku △△-cho," with a space in between, to generate a single string, "a line has formed ramen shop Shinjuku-ku △△-cho." This allows the control unit 17 to generate a search term that can be entered into the search box of a search engine.

[0044] The following describes an example of how the above-mentioned search words are registered, with reference to Figure 3. Figure 3 is a map of the area around the utterance location displayed on the touch panel display 14.

[0045] In this embodiment, the control unit 17 generates a search word and then registers the search word in association with the utterance location on the map. Specifically, as shown in Figure 3, the control unit 17 displays the utterance location UP superimposed on the map and displays the search word ("Ramen shop with a line, Shinjuku-ku △△-cho") in a speech bubble to correspond to the utterance location UP.

[0046] For example, if a passenger in vehicle M touches the speech bubble displayed on the touch panel display 14, the control unit 17 connects to a search site such as Google® or Yahoo!®. The control unit 17 then enters a search term into the search box of the launched search site and performs the search.

[0047] On the touch panel display 14, information such as web pages and images that are highly relevant to the search term are displayed preferentially as a result of the search. This allows passengers of vehicle M to obtain information that is directly or indirectly related to the subject TA that they are interested in.

[0048] In other words, in this embodiment, the driver of vehicle M who asks the question, "What kind of ramen shop is that with the long line?" can, by touching the speech bubble displayed on the touch panel display 14 ("Ramen shop with a long line, Shinjuku-ku △△-cho"), find a website that introduces popular ramen shops in Shinjuku-ku △△-cho, and from there obtain information about the target ramen shop TA, such as its name and the type of ramen it serves.

[0049] The specific operation of the in-vehicle device 10 in this embodiment will be described below. Figure 4 is a flowchart of the search word generation routine RT1 executed in the control unit 17 of the in-vehicle device 10.

[0050] In the following explanation, we will assume, as explained above, that the user has entered the following voice message: "What kind of ramen shop is that with the long line?"

[0051] The control unit 17 starts the search word generation routine RT1, for example, when power is turned on to the in-vehicle device 10. The search word generation routine RT1 is repeatedly executed by the control unit 17 as long as the in-vehicle device 10 is powered on.

[0052] First, the control unit 17 determines whether or not it has received a search word generation operation (step S101). For example, the control unit 17 determines that an operation to start the search word generation process has been performed when the user calls out the wake word for the in-vehicle device 10 or makes a predetermined voice input such as "Make a note of this!". If the control unit 17 determines that it has not received a search word generation operation (step S101: NO), it terminates the search word generation routine RT1.

[0053] When the control unit 17 determines that it has received a search word generation operation (step S101: YES), it accepts voice input from a passenger in vehicle M (step S102). For example, as described above, the control unit 17 acquires voice from the driver of vehicle M via microphone 12 while the vehicle M is in motion. In step S102, if there is no voice input from a passenger in vehicle M within a certain period of time, the search word generation routine RT1 may be terminated.

[0054] After step S102, the control unit 17 uses the speech recognition DB22A of the mass storage device 22 to perform speech recognition on the input speech and convert the speech into text (step S103). In this embodiment, the control unit 17 converts the speech input by the user, "What kind of restaurant is that ramen shop with the long line?", into text consisting of strings.

[0055] After step S103, the control unit 17 extracts common nouns from the text resulting from speech recognition (step S104). In this embodiment, the control unit 17 extracts "ramen shop," which represents a common noun.

[0056] After step S104, the control unit 17 extracts words from the speech recognition result text that indicate attributive modifiers, attributive phrases, or attributive clauses relating to common nouns (step S105). In this embodiment, the control unit 17 extracts the words "there is a line" which indicates an attributive clause relating to "ramen shop".

[0057] After step S105, the control unit 17 generates a first word using the common noun extracted in steps S104 and S105 and the words indicating the attributive modifier, attributive phrase, or attributive clause related to the common noun (step S106). In this embodiment, as described above, the control unit 17 uses "ramen shop" which indicates the common noun and "a line has formed" which indicates the attributive clause related to "ramen shop" as the first word.

[0058] After generating the first word in step S106, the control unit 17 identifies a place name based on the utterance location and map information, and generates a second word based on the identified place name (step S107). The control unit 17 generates "Shinjuku-ku △△-cho" as the second word.

[0059] The control unit 17 generates a search word as a string using the first word generated in step S106 and the second word generated in step S107, and registers it in association with the utterance position (step S108).

[0060] Specifically, as described above, the control unit 17 combines the first words, "a line has formed" and "ramen shop," with the second word, "Shinjuku-ku △△-cho," with a space in between, to generate the first search word, "a line has formed ramen shop Shinjuku-ku △△-cho."

[0061] Then, after generating a search word, the control unit 17 registers the search word in association with the utterance location on the map. Specifically, as shown on the touch panel display 14 in Figure 3, the control unit 17 displays a speech bubble with "A ramen shop with a long line, Shinjuku-ku △△-cho" associated with the utterance location UP. After step S108, the control unit 17 terminates the search word generation routine RT1.

[0062] Figure 5 is a flowchart showing the second word generation subroutine RT2 executed by the control unit 17 in step S107 described above.

[0063] First, the control unit 17 acquires the speech location of the passenger in vehicle M and map data (step S201). Specifically, as described above, the control unit 17 acquires the speech location based on the GPS signal received from the GPS receiver 13 when voice input is made by speech. The control unit 17 also acquires map information from the map information DB 22B of the large-capacity storage device 22.

[0064] After step S201, the control unit 17 identifies the address of the utterance location based on the utterance location and map information acquired in step S201 (step S202). Specifically, as described above, the control unit 17 identifies the address of the utterance location on the map, which is "Shinjuku-ku △△-cho ○-chome ×-ban □-go".

[0065] The control unit 17 identifies the place name based on the address identified in step S202 (step S203). That is, the control unit 17 identifies "Shinjuku-ku △△-cho ○-chome ×-ban □-go," which is the address of the utterance location, as the place name.

[0066] The control unit 17 generates a second word based on the identified place name (step S204). In this embodiment, as described above, the control unit 17 uses "Shinjuku-ku △△-cho," which is the town name part of the address of the utterance location identified in step S203, as the second word. After step S204, the control unit 17 terminates the second word generation subroutine RT2.

[0067] According to this embodiment, the search word generation routine RT1 and the second word generation subroutine RT2 described above automatically generate search words based on voice input from the passengers of vehicle M. Therefore, when passengers of vehicle M want to search for a subject TA that interests them, they can avoid the trouble of thinking of search words themselves and entering them into the search box.

[0068] The generated search words are then displayed on the touch panel display 14 in a manner that corresponds to the location of the spoken word on the map. Therefore, passengers of vehicle M can easily search for the search words at any time and obtain information that is highly relevant to the search words. Accordingly, according to this embodiment, users of the in-vehicle device 10 can easily obtain information about objects they are interested in while traveling, at any time they like.

[0069] In this embodiment, the control unit 17, when generating the first word (step S106), uses "ramen shop," which represents the extracted common noun, and "a line has formed," which represents the attributive clause related to the common noun, as the first word. However, this is not limited to this, and the extracted word may be modified as appropriate.

[0070] For example, the control unit 17 may use the attributive clause "there is a line" relating to "ramen shop" simply as "line." In other words, the control unit 17 may use "ramen shop" and "line" as the first word.

[0071] Furthermore, in this embodiment, the control unit 17, when generating the second word (step S204), uses the town name from the address of the identified utterance location as the second word, but this is not limited to this and may be changed as appropriate.

[0072] For example, if the address at the utterance location is "Shinjuku-ku △△-cho ○-chome ×-ban □-go", the control unit 17 may use "Shinjuku-ku △△-cho ○-chome" as the second word, or it may use "Shinjuku-ku △△-cho ○-chome ×-ban" as the second word. In other words, the control unit 17 may freely change the degree to which the address adopted for the second word is narrowed down.

[0073] In this embodiment, the control unit 17 extracts common nouns from the user's utterances, but is not limited to this; it may also extract nouns that do not uniquely identify the subject. Such nouns may include, for example, the trade name, store name, or brand name of a company with multiple stores.

[0074] Specifically, for example, if a user says, "What kind of restaurant is that 'Ramen XX' with the long line?", the name of the ramen shop, "Ramen XX," can be extracted and combined with the address of the location where the utterance occurred to create a search term. For example, in the case of a ramen shop with multiple locations, the name of the ramen shop alone may not be enough to identify a single shop, so generating the above search term is effective.

[0075] In this embodiment, the control unit 17 displays registered search words on the touch panel display 14 in a bubble corresponding to the utterance location on the map, and the user performs the search by touching the bubble display; however, this is not limited to this embodiment.

[0076] For example, as long as the search term is stored in association with the utterance location on the map, the control unit 17 does not need to display the search term itself. For example, the speech bubble display described above may include only a part of the search term, such as "there is a line" or "ramen shop," or it may include the entire text of the user's voice input, or it may include the date and time of the utterance.

[0077] Alternatively, for example, the control unit 17 may launch a search site immediately after generating the search term in step S108, and then enter the generated search term into the search box of the search site to perform a search. This allows the user to automatically generate the search term and perform the search operation when voice input is made, thus saving the user the trouble of performing the search operation themselves.

[0078] Furthermore, the control unit 17 may output the search results based on the search term as audio via the speaker 15. In other words, the search results may be communicated to the user without the user having to look at the touch panel display 14.

[0079] Furthermore, the speech recognition of the input voice performed by the control unit 17 may be performed by AI. For example, the voice input by the user's utterance may be input to an AI having a learning model trained on a speech dictionary or language model, and text consisting of strings as its output may be obtained.

[0080] In this embodiment, the above-described search word generation process is shown to be performed by an in-vehicle device 10 mounted on a vehicle M, which is a mobile object. However, the embodiment is not limited to this, as long as the device configuration is capable of performing the process. For example, each configuration of the in-vehicle device 10 may be applied to a portable communication terminal that can move with the vehicle, such as a smartphone or tablet.

[0081] Furthermore, for example, the search term generation process described above may be performed by a server device capable of communicating with the in-vehicle device 10 instead of the in-vehicle device 10, and the in-vehicle device 10 may receive the results of the process from the server device.

[0082] Specifically, for example, the in-vehicle device 10 may transmit the voice and utterance location input to the microphone 12 from the passenger of vehicle M to the server device, and the server device may perform the search word generation process. The in-vehicle device 10 may then display information on the touch panel display 14 of the in-vehicle device 10 that indicates the search word registered in correspondence with the utterance location on the map, which is transmitted from the server device.

[0083] Furthermore, the information transmitted from the server device, which indicates registered search words associated with the speech location on the map, may be configured to be viewable on portable communication devices that can be moved with the vehicle, such as smartphones and tablet devices.

[0084] Alternatively, the user may input voice commands into a portable communication device that can be moved with the vehicle, such as a smartphone or tablet, and transmit the voice and utterance location from the communication device to a server device, where the server device will perform the search word generation process. The in-vehicle device 10 may then receive information indicating the registered search words associated with the utterance location on the map and display this information on the touch panel display 14. [Examples]

[0085] Next, Example 2 will be described. In Example 2, some of the configuration of the in-vehicle device 10 and the method by which the control unit 17 identifies place names in response to user utterances differ from Example 1, but all other aspects are the same as in Example 1.

[0086] Figure 6 shows the configuration of the front seat area of ​​vehicle M according to Embodiment 2 and the view in front of vehicle M. In Figure 6, a cylindrical structure extending vertically to the right front of the direction of travel of vehicle M is visible from inside vehicle M while the vehicle is moving.

[0087] In Figure 6, the driver of vehicle M, curious about what kind of structure is visible from inside vehicle M, asks microphone 12, "What is that chimney-like thing on the front right?"

[0088] In this embodiment, the in-vehicle device 10, similar to Embodiment 1, extracts common nouns and words indicating attributive modifiers, attributive phrases, or attributive clauses related to the common nouns from the input voice from the passenger of the vehicle M, and generates a first word based on the common nouns and the words indicating attributive modifiers, attributive phrases, or attributive clauses.

[0089] Furthermore, in this embodiment, the in-vehicle device 10 identifies a place name based on the extracted attributive modifier, attributive phrase, or attributive clause, the utterance location described above, and map data, and generates a second word based on that place name. The in-vehicle device 10 combines the generated first word and the second word to generate a search word and registers it in association with the utterance location on the map.

[0090] Figure 7 is a block diagram showing an example of the configuration of the in-vehicle device 10 according to Example 2. In Figure 7, the differences from the configuration of the in-vehicle device 10 shown in Example 1 will be explained.

[0091] In this embodiment, the large-capacity storage device 22 of the in-vehicle device 10 has a range setting DB22C. The range setting DB22C is a database used by the control unit 17 to identify place names when generating the second word described above.

[0092] In this embodiment, the control unit 17, while the vehicle M is moving, considers the direction in front of the vehicle M (i.e., the direction of travel) as the direction the user is facing, based on the direction of travel of the vehicle M at the point of speech and the map data stored in the map information DB22B, and acquires it as a reference direction. In other words, the control unit 17 is a direction acquisition unit that acquires the direction the user is facing as a reference direction.

[0093] The following describes an example of the search word generation process performed by the control unit 17 in this embodiment. In the following description, we will assume that the user has input the following voice message, "What is that chimney-like thing on the front right?" as described above.

[0094] [Extracting text] First, let's explain the text extraction. Similar to Embodiment 1, the control unit 17 extracts common nouns and words indicating attributive modifiers, attributive phrases, or attributive clauses related to those common nouns from the text converted by speech recognition.

[0095] Specifically, the control unit 17 extracts "chimney," which is a common noun, and "front right," which is an attributive modifier related to "chimney," from the string "What is that thing that looks like a chimney on the front right?" which has been converted to text by speech recognition. In this embodiment, the expression "thing" in "What is that thing that looks like a chimney on the front right?" is a formal noun, so the control unit 17 extracts "chimney," which is a common noun, without extracting "thing."

[0096] [Generating the first word] Next, the generation of the first word will be explained. In this embodiment, the control unit 17 identifies the word "right front," which indicates an attributive modifier related to "chimney," as an expression indicating direction. In this embodiment, when an expression indicating direction is identified from a word indicating an attributive modifier, attributive phrase, or attributive clause related to a common noun, the control unit 17 generates the first word without including that expression. That is, in this embodiment, the control unit 17 uses only the common noun "chimney" as the first word.

[0097] [Generating the second word] Next, the generation of the second word will be explained using Figure 8. Figure 8 is a diagram showing an example of the method of identifying a place name by the control unit 17 in Embodiment 2. In Figure 8, the boundary of the town indicated by town name 1 is shown by a dashed line. Also in Figure 8, the position of vehicle M is set as the utterance position where the passenger of vehicle M spoke. At the position of vehicle M, a speech bubble is displayed showing the words indicating the voice input by the passenger of vehicle M.

[0098] In this embodiment, as shown in Figure 8, the control unit 17 sets a triangular area R (two-dot dashed line in the figure) on the map starting from the utterance location, and identifies a place name that is partially or entirely contained within this area R. Then, the control unit 17 generates a second word based on the identified place name. In this embodiment, the area R is set using a distance range D determined based on the attributes of what the common noun represents, and an angle range θ determined based on the expression indicating direction.

[0099] Here, the distance range D and the angular range θ will be explained with reference to Figures 9 and 10.

[0100] Figure 9 shows examples of common nouns classified by genre in this embodiment, and the distance range TB1 that indicates the corresponding distance range D. The control unit 17 refers to the distance range TB1 and identifies the distance range D that corresponds to the attribute of the common noun.

[0101] This distance range D may be determined, for example, by how far away the object indicated by the common noun can be seen. For example, if the object indicated by the common noun is a building, the taller it is, the further away it can be seen from, and therefore the larger the distance range D will be.

[0102] In distance range TB1, for example, if the common noun in the input voice refers to movable or immovable property such as "shop," "stall," or "mobile vending machine," which are several meters tall, then distance range D is 100m.

[0103] Furthermore, within the distance range TB1, for example, if the common noun in the input voice refers to movable or immovable property such as "building," "chimney," or "ferry," which are several tens to several hundreds of meters tall, then the distance range D is 1 km.

[0104] Furthermore, within the distance range TB1, for example, if the common noun in the input voice refers to an event held indoors or outdoors, such as an "exhibition," "concert," or "festival," then the distance range D is 300m.

[0105] Furthermore, within the distance range TB1, for example, if the common noun in the input voice refers to an event that includes objects visible from a distance, such as a "fireworks display" or advertising balloons, then the distance range D is 3km.

[0106] In this embodiment, since the common noun in the input voice is "chimney," the control unit 17 refers to the distance range TB1 and determines that the distance range D corresponding to "chimney" is 1 km.

[0107] Furthermore, since things that can be seen from a distance are likely to be spoken of before approaching them, a minimum distance for distance range D may be set. For example, if the common noun refers to a building, chimney, ferry, etc., the minimum distance may be set to 1 km and the distance range D to be from 1 km to 3 km.

[0108] Figure 10 shows an example of a representation indicating direction in this embodiment and the corresponding angular range TB2. The control unit 17 refers to the angular range TB2 to determine the angular range θ corresponding to the representation indicating direction.

[0109] Within the angular range TB2, for example, if the control unit 17 identifies an expression indicating the direction relative to its own reference direction, such as "the previous...", in the input voice, it determines that the angular range θ is a range from 45° to the left (-45°) to 45° to the right of the reference direction it is facing, with the reference direction being the direction it is facing (0°).

[0110] Furthermore, within the angular range TB2, for example, if the control unit 17 identifies an expression indicating the direction relative to its own reference direction, such as "left...", as a word modifying a common noun in the input voice, it determines that the angular range θ is from 225° to 315°, with its own direction as the reference (0°).

[0111] In this embodiment, as described above, the control unit 17 identifies the phrase "front right," which is an attributive modifier related to the common noun "chimney," as an expression indicating the direction relative to the reference direction in which the occupant of vehicle M is facing. Based on this, the control unit 17 refers to the angle range TB2 and identifies the angle range θ corresponding to "front right" as being in the range from 0° to 45°, with the direction it is facing as the reference (0°).

[0112] Refer to Figure 8 again. As described above, when the control unit 17 receives the input voice "What is that chimney-like thing on the right front?", it identifies the distance range D as 1 km based on the distance range TB1, using the word "chimney," which is a common noun. The control unit 17 also identifies the angle range θ as being between 0° and 45° based on the angle range TB2, using the word "on the right front," which is an attributive phrase related to the common noun.

[0113] The control unit 17 forms a range R on the map based on the specified distance range D and angle range θ. Specifically, the control unit 17 defines a first side as an extension of the distance range D from the utterance position in the direction it is facing, a second side as an extension of the first side from the utterance position by the angle range θ, and forms a triangular range R by connecting the endpoint of the first side and the endpoint of the second side with a straight line.

[0114] The control unit 17 then identifies place names that are partially or entirely contained within the range R. Specifically, as shown in Figure 8, the control unit 17 identifies "Shinjuku-ku △△-cho" and "Shinjuku-ku ×○-cho" as place names.

[0115] In this embodiment, the control unit 17 uses the identified place name as the second word. That is, in this embodiment, the control unit 17 uses "Shinjuku-ku △△-cho" and "Shinjuku-ku ×○-cho" as the second word.

[0116] [Generating search terms] Next, the generation of search words will be explained. Similar to Embodiment 1, the control unit 17 generates search words by combining the first and second words that were generated.

[0117] The control unit 17 generates a search term as an OR search + AND search if there are multiple second words generated. Specifically, the control unit 17 generates a search term "(Shinjuku-ku △△-cho OR Shinjuku-ku ×○-cho) chimney" which includes the second words "Shinjuku-ku △△-cho" and "Shinjuku-ku ×○-cho" respectively, and the first word "chimney".

[0118] The search words generated by the control unit 17 are registered in association with the speech location on the map, similar to the first embodiment. A passenger in vehicle M can obtain the search results using the search words by touching the speech bubble displaying "(Shinjuku-ku △△-cho OR Shinjuku-ku ×○-cho) chimney" on the touch panel display 14, for example, similar to the first embodiment.

[0119] The second word generation process performed in the control unit 17 of the in-vehicle device 10 in this embodiment will be described below. Figure 11 is a flowchart of the second word generation subroutine RT3 performed in the control unit 17 of the in-vehicle device 10. Step S201, which performs the same processing as in Embodiment 1, will not be explained.

[0120] After acquiring the utterance position and map data in step S201, the control unit 17 determines the distance range D based on the attributes of the common noun described above, and determines the angle range θ based on the expression indicating the direction described above (step S301).

[0121] After step S301, the control unit 17 forms a range R on the map using the specified distance range D and angle range θ (step S302). Specifically, as shown in Figure 8, a triangular range R is formed starting from the utterance position.

[0122] After step S302, the control unit 17 identifies place names within the range R (step S303). In other words, in this embodiment, the control unit 17 identifies "Shinjuku-ku △△-cho" and "Shinjuku-ku ×○-cho" within the range R as place names.

[0123] The control unit 17 generates a second word based on the place name identified in step S303 (step S304). In this embodiment, the control unit 17 uses "Shinjuku-ku △△-cho" and "Shinjuku-ku ×○-cho" as the second words. After step S304, the control unit 17 terminates the second word generation subroutine RT3.

[0124] According to this embodiment, similar to Embodiment 1, the generated search words are registered in a manner that associates them with the utterance location on the map, so that users such as passengers in vehicle M can perform searches on the search words at any time.

[0125] Furthermore, according to this embodiment, the control unit 17 sets a range for identifying a place name using a distance range based on the attributes of the common noun in the input voice and an angle range based on the expression indicating the direction relative to the reference direction it is facing in the input voice.

[0126] As a result, the control unit 17 can generate search terms that target the region where the target TA is expected to be located, for example, when the user is far away from the target TA of their interest. Therefore, the user can obtain information that is highly relevant to the search terms, that is, information that is directly or indirectly related to the target TA.

[0127] Therefore, according to this embodiment, users can easily obtain information about objects that interest them while they are on the move, at any time they like.

[0128] In this embodiment, the distance range TB1 and angle range TB2 shown are merely examples, and their ranges may be changed as appropriate. For example, the distance range may be a fixed distance range regardless of the attributes of the noun, and the control unit 17 may form range R based on this fixed distance range and angle range TB2.

[0129] Furthermore, in this embodiment, the control unit 17 forms a triangular area R on the map, but its shape is not limited to this. For example, when the control unit 17 defines the first and second sides as described above, it may connect the endpoint of the first side and the endpoint of the second side with an outwardly convex curve. In other words, the area R may take the form of a sector.

[0130] In this embodiment, as in Embodiment 1, a communication terminal such as a smartphone or tablet may be used. In this case, the control unit 17 acquires the direction in which the user is moving as the direction the user is facing.

[0131] In this embodiment, the control unit 17 generates a search term "(Shinjuku-ku △△-cho OR Shinjuku-ku ×○-cho) chimney" as an OR search + AND search, but the form of the search term is not limited to this. For example, the control unit 17 may simply generate the search terms "Shinjuku-ku △△-cho chimney" and "Shinjuku-ku ×○-cho chimney" and register them at the utterance position, allowing the user to select the one they prefer. [Examples]

[0132] Next, Example 3 will be described. In Example 3, the method by which the control unit 17 identifies place names in response to user utterances differs from that of Example 2, but all other aspects are the same as in Example 2.

[0133] Figure 12 shows the configuration of the front seat area of ​​vehicle M according to Embodiment 3 and the view in front of vehicle M. In Figure 12, while vehicle M is moving, the ○× Tower is visible to the left front of the direction of travel of vehicle M, and a building with a cloud-like object is visible to the right of the ○× Tower.

[0134] In Figure 12, the driver of vehicle M, curious about what kind of building it is with the cloud-like object, asks microphone 12, "What is that building to the right of the ○× Tower with the cloud-like object?"

[0135] In this embodiment, the in-vehicle device 10, similar to Embodiment 2, extracts common nouns and words indicating attributive modifiers, attributive phrases, or attributive clauses related to the common nouns from the input voice from the passenger of the vehicle M, and generates a first word based on the common nouns and the words indicating attributive modifiers, attributive phrases, or attributive clauses.

[0136] In this embodiment, the in-vehicle device 10 identifies a place name based on the extracted attributive modifier, attributive phrase, or attributive clause, the utterance location described above, and map data, and generates a second word based on that place name. Similar to Embodiment 2, the in-vehicle device 10 combines the generated first word and the second word to generate a search word and registers it in association with the utterance location on the map.

[0137] The following describes an example of the search word generation process performed by the control unit 17 in this embodiment. In the following description, we will assume, as described above, that the user has input the following voice message: "What is the building with the cloud-like object to the right of the ○× Tower?"

[0138] [Extracting text] First, let's explain the text extraction. Similar to Embodiment 2, the control unit 17 extracts common nouns and words indicating attributive modifiers, attributive phrases, or attributive clauses related to those common nouns from the text converted by speech recognition.

[0139] Specifically, the control unit 17 extracts the common noun "building" and the attributive phrase "the building with the cloud-like object to the right of the ○× Tower" from the string "What is the building with the cloud-like object to the right of the ○× Tower?" which has been converted into text by speech recognition.

[0140] [Generating the first word] Next, the generation of the first word will be explained. In this embodiment, the control unit 17 identifies the attributive clause "to the right of ○× Tower" relating to "building" as an expression that includes the proper noun "○× Tower" and indicates the direction relative to the reference direction, where the direction from the utterance position toward what the proper noun indicates (○× Tower) is used as the reference direction.

[0141] In this embodiment, the control unit 17 generates a first word without including expressions indicating direction relative to a proper noun and the direction from the utterance position toward the proper noun, when such expressions are identified from words indicating attributive modifiers, attributive phrases, or attributive clauses relating to a common noun. That is, in this embodiment, the control unit 17 uses "building," which indicates a common noun, and "there is a cloud-like object," which indicates an attributive clause relating to "building," as the first word.

[0142] [Generating the second word] Next, the generation of the second word will be explained using Figure 13. Figure 13 is a diagram showing an example of the method of identifying a place name by the control unit 17 in Embodiment 3. In Figure 13, the boundary of the town indicated by town name 1 is shown by a dashed line. Also, in Figure 13, the position of vehicle M is set as the utterance position where the passenger of vehicle M spoke.

[0143] In this embodiment, the control unit 17 identifies the direction relative to what the proper noun indicates and the direction from the utterance position toward what the proper noun indicates, based on the words indicating attributive modifiers, attributive phrases, or attributive clauses relating to common nouns in the input speech.

[0144] In this embodiment, proper nouns refer to landmarks that serve as reference points for the occupants of vehicle M to indicate their positional relationship with the target TA. In this embodiment, the landmarks are assumed to have their names and locations pre-registered in map information.

[0145] In this embodiment, the control unit 17, similar to Embodiment 2, sets a triangular area R (a dashed line in the figure) enclosed by the utterance position, and identifies place names on the map that are partially or entirely contained within area R. Then, the control unit 17 generates a second word based on the identified place names.

[0146] In this embodiment, the predetermined range R is set using a distance range D determined based on the attributes of the common noun, and an angular range θ determined based on an expression indicating the direction from the utterance location toward the proper noun (landmark). The distance range D is determined in the same manner as in Embodiment 2.

[0147] Here, we will explain the angular range θ with reference to Figure 14. Figure 14 shows an example of an expression indicating the direction relative to the direction from the utterance position toward the landmark (hereinafter also referred to as LM), and the angular range TB3 which shows the corresponding angular range θ.

[0148] In the angular range TB3, for example, if words such as "behind LM..." or "beyond LM..." are identified from words modifying common nouns in the input speech, the angular range θ is defined as a range from 45° to the left (-45°) to 45° to the right of the reference point (0°) which is the direction from the utterance position toward LM, with clockwise rotation being positive.

[0149] Furthermore, in the angular range TB3, for example, if the phrase "to the left of LM..." is identified from a word modifying a common noun in the input speech, the angular range θ is from 315° to 360°, with the direction from the utterance position toward LM as the reference (0°).

[0150] In this embodiment, the control unit 17 identifies the phrase "to the right of ○× Tower" as an expression indicating the direction relative to the direction from the utterance position toward "○× Tower". Based on this, the control unit 17 refers to the angular range TB3 and identifies the angular range θ corresponding to "to the right of LM" as being in the range from 0° to 45°, with the direction from the utterance position toward LM as the reference (0°).

[0151] Refer to Figure 13 again. When the control unit 17 receives the input voice message, "What is the building with the cloud-like object to the right of the ○× Tower?", it identifies the distance range D as 1 km based on the distance range TB1 shown in Figure 9, using the word "building," which is a common noun. The control unit 17 also identifies the angle range θ as being between 0° and 45° based on the angle range TB3, using the word "to the right of the ○× Tower," which is an attributive clause modifying the common noun.

[0152] The control unit 17 forms the above-described range R on the map based on the identified distance range D and angle range θ. Then, the control unit 17 identifies place names within range R on the map. Specifically, as shown in Figure 13, the control unit 17 identifies "Shinjuku-ku △△-cho", "Shinjuku-ku ×○-cho", and "Shinjuku-ku △□-cho" as place names.

[0153] In this embodiment, the control unit 17 uses the specified place name as the second word, similar to the second embodiment. That is, in this embodiment, the control unit 17 uses "Shinjuku-ku △△-cho", "Shinjuku-ku ×○-cho", and "Shinjuku-ku △□-cho" as the second word.

[0154] [Generating search terms] Next, the generation of search words will be explained. Similar to Embodiment 2, the control unit 17 combines the first and second words that have been generated to generate a search word as an OR search + AND search.

[0155] Specifically, the control unit 17 generates a search term "(Shinjuku-ku △△-cho OR Shinjuku-ku ×○-cho OR Shinjuku-ku △□-cho) cloud-like object building" which includes the second word "Shinjuku-ku △△-cho", "Shinjuku-ku ×○-cho", or "Shinjuku-ku △□-cho", and the first words "there is a cloud-like object" and "building". The search term generated by the control unit 17 is registered in association with the utterance location on the map, as in Examples 1 and 2. The second word generation process performed in the control unit 17 of the in-vehicle device 10 in this embodiment will be described below. Figure 15 is a flowchart of the second word generation subroutine RT4 performed in the control unit 17 of the in-vehicle device 10. Step S201, which performs the same processing as in Embodiment 1, will not be explained.

[0156] After acquiring the utterance position and map data in step S201, the control unit 17 identifies the location of the object (landmark) indicated by the proper noun based on the map data (step S401). Then, after step S401, the control unit 17 identifies the direction from the utterance position toward the object indicated by the proper noun as the reference direction based on the location of the object indicated by the proper noun (step S402).

[0157] After step S402, the control unit 17 determines a distance range D based on the attributes of the thing indicated by the common noun, and determines an angular range θ based on an expression indicating the direction relative to the reference direction from the utterance position toward the thing indicated by the proper noun (step S403).

[0158] After step S403, the control unit 17 forms a range R on the map using the specified distance range D and angle range θ (step S404). Specifically, as shown in Figure 13, a triangular range R is formed starting from the utterance position.

[0159] After step S404, the control unit 17 identifies place names within the range R (step S405). In other words, in this embodiment, the control unit 17 identifies "Shinjuku-ku △△-cho", "Shinjuku-ku ×○-cho", and "Shinjuku-ku △□-cho" within the range R as place names.

[0160] The control unit 17 generates a second word based on the place name identified in step S405 (step S406). In this embodiment, the control unit 17 uses "Shinjuku-ku △△-cho", "Shinjuku-ku ×○-cho", and "Shinjuku-ku △□-cho" as the second words. After step S406, the control unit 17 terminates the second word generation subroutine RT4.

[0161] According to this embodiment, similar to Embodiment 2, the generated search words are registered in a manner that associates them with the utterance location on the map, so that users such as passengers in vehicle M can search for the search words at any time.

[0162] Furthermore, according to this embodiment, the control unit 17 can identify one or more place names for second word generation by identifying what is indicated by a proper noun as a landmark from the input voice uttered by the user, and by identifying an expression indicating the direction from the utterance location toward the landmark.

[0163] Therefore, the control unit 17 can, for example, generate search terms that target the area where the target TA is expected to be located, while referring to landmarks, especially when there is a distance between the user and the target TA of their interest.

[0164] Therefore, users can obtain information that is highly relevant to their search terms, that is, information that is directly or indirectly related to the target TA. Accordingly, according to this embodiment, users can easily obtain information about a subject they are interested in while on the go, at any time they like.

[0165] In this embodiment, when the control unit 17 identifies an expression indicating the direction relative to what the proper noun represents from the input voice and the direction from the utterance position toward what the proper noun represents, it does not include that expression in the first word, but it is not limited to this.

[0166] For example, the control unit 17 may include what a proper noun represents in the first word. Specifically, the control unit 17 may include "○× Tower" as a proper noun in the first word and generate a search word as an OR search + AND search: "(Shinjuku-ku △△-cho OR Shinjuku-ku ×○-cho OR Shinjuku-ku △□-cho) Building with a cloud-like object ○× Tower". [Examples]

[0167] Next, Example 4 will be described. In Example 4, the control unit 17 identifies the second word, which differs from Example 1. Other aspects, such as the configuration of each component of the in-vehicle device 10 and the method for generating the first word, are the same as in Example 1.

[0168] In this embodiment, the control unit 17 identifies a feature in the vicinity of a speech location based on the speech location input by the passenger of the vehicle M and map data. In this embodiment, the control unit 17 generates a second word based on the identified feature.

[0169] The following describes an example of the search word generation process in this embodiment. Note that the text extraction and the first word generation process are the same as in Embodiment 1 and will therefore be omitted. In the following explanation, as in Embodiment 1, we will assume that the user has input the voice message, "What kind of ramen shop is that with the long line?"

[0170] [Generating the second word] The following describes the differences between this embodiment and Embodiment 1 regarding the generation of the second word. Figure 16 shows an example of the method of identifying features by the control unit 17 in Embodiment 3. In Figure 16, the boundary of the town indicated by town name 1 is shown by a dashed line. Also, in Figure 16, the position of vehicle M is set as the utterance position where the passenger of vehicle M spoke.

[0171] The control unit 17 identifies a feature located around the location where the passenger of vehicle M said, "What kind of ramen shop is that with the long line?", based on the location of the utterance and the map information stored in the map information DB 22B.

[0172] Specifically, the control unit 17 identifies the names of facilities related to transportation infrastructure, such as the nearest train station, bus stop, railway line, or road line, located around the speech location, based on the speech location and map information. In this embodiment, as shown in Figure 16, the control unit 17 identifies "○△ Station," which is the nearest train station to the speech location. In this embodiment, the control unit 17 uses "○△ Station," which is the nearest train station to the speech location, as the second word.

[0173] [Generating search terms] The control unit 17 combines the first generated words, "there is a line" and "ramen shop," with the second word, "○△ Station," with a space in between, to generate the first search word, "there is a line ramen shop ○△ Station." Then, the control unit 17 registers the generated search word in association with the utterance position.

[0174] The second word generation process performed in the control unit 17 of the in-vehicle device 10 in this embodiment will be described below. Figure 17 is a flowchart of the second word generation subroutine RT5 performed in the control unit 17 of the in-vehicle device 10. Step S201, which performs the same processing as in Embodiment 1, will not be explained.

[0175] After acquiring the speech location and map data (step S201), the control unit 17 identifies a feature located around the speech location based on the speech location and map information (step S501). Specifically, as described above, the control unit 17 identifies "○△ Station," which is the nearest station to the speech location on the map.

[0176] The control unit 17 generates a second word using the feature identified in step S501 (step S502). In this embodiment, as described above, the control unit 17 uses "○△ Station," which is a word indicating a feature around the utterance location identified in step S501, as the second word. After step S502, the control unit 17 terminates the second word generation subroutine RT5.

[0177] According to this embodiment, similar to Embodiment 1, the generated search words are registered in a manner that associates them with the utterance location on the map, so that users such as passengers in vehicle M can perform searches on the search words at any time.

[0178] Furthermore, according to this embodiment, the control unit 17 identifies a feature located around the utterance position, uses that feature as a second word, and generates a search word using that second word. Therefore, according to this embodiment, for example, if there are articles on the internet that introduce shops in relation to transportation infrastructure, such as "There is always a line at the ramen shop near ○△ Station," such articles can be efficiently searched.

[0179] In this embodiment, the control unit 17 identifies features located around the utterance position, but the number of features may be multiple. For example, the control unit 17 may identify the aforementioned "○△ Station" and the name of the train line that passes through "○△ Station," and both may be used as the second word. In this case, as explained in Embodiment 2, the search word may be generated using OR search + AND search.

[0180] Furthermore, the control unit 17 may pre-identify multiple features located around the utterance position and designate the feature closest to the utterance position as the second word. Additionally, a priority order for designating each feature as the second word may be established based on its attributes; for example, the name of a nearby station may be given the highest priority as the second word.

[0181] In this embodiment, the control unit 17 identifies a feature in the vicinity of the utterance location, but is not limited to this; it may also identify a road route at the utterance location as a feature. For example, when vehicle M is traveling on Loop Route 7, and a passenger in vehicle M says, "What kind of ramen shop is that with the long line?", the control unit 17 may identify "Loop Route 7" as the second word.

[0182] The control unit 17 may then create a search term such as "Ramen shop with a long line, Loop Route 7". This allows the user to obtain information such as online articles introducing ramen shops located along Loop Route 7.

[0183] Furthermore, the geographical features identified by the control unit 17 may include not only facilities related to the transportation infrastructure described above, but also natural features such as lakes and coastlines, and these may be included in the search terms. This allows for efficient searching of articles, for example, online such as "A famous restaurant along the XX coast..." [Examples]

[0184] Next, Example 5 will be described. In Example 5, the control unit 17 identifies the second word, which differs from Example 2. Other aspects, such as the configuration of each component of the in-vehicle device 10 and the method for generating the first word, are the same as in Example 2.

[0185] In this embodiment, the control unit 17 identifies features around the location of a utterance based on the voice input received from the utterance of a passenger in the vehicle M, the location of the utterance, and map data. In this embodiment, the control unit 17 generates a second word based on the identified features.

[0186] The following describes an example of the search word generation process in this embodiment. Note that the text extraction and the first word generation process are the same as in Embodiment 2 and will therefore be omitted. In the following explanation, as in Embodiment 2, we will assume that the user has input the voice message, "What is that chimney-like thing on the front right?"

[0187] [Generating the second word] The following describes the differences between this embodiment and Embodiment 2 regarding the generation of the second word.

[0188] Figure 18 shows an example of the method for identifying features by the control unit 17 in Embodiment 3. In Figure 18, the boundary of the town indicated by town name 1 is shown by a dashed line. Also, in Figure 18, the position of vehicle M is set as the utterance position where the passenger of vehicle M spoke.

[0189] In this embodiment, the control unit 17, similar to Embodiment 2, sets a triangular area R (two-dot dashed line in the figure) enclosed by the utterance position and identifies features within area R on the map. Then, the control unit 17 generates a second word based on the identified features. The predetermined area R is set, similar to Embodiment 2, using a distance area D determined based on the attributes of what the common noun represents and an angle area θ determined based on the direction expression described above.

[0190] Similar to Embodiment 2, when the control unit 17 receives the input voice "What's that chimney-like thing on the front right?", it identifies the distance range D as 1 km based on the distance range TB1, using the common noun "chimney". The control unit 17 also identifies the angle range θ as being between 0° and 45° based on the angle range TB2, using the attributive phrase "front right" which modifies the common noun.

[0191] The control unit 17 sets the range R described above based on the specified distance range D and angle range θ. Then, the control unit 17 identifies features within the set range R on the map. Specifically, as shown in Figure 18, the control unit 17 identifies "National Highway No. ▽" which is a road route within range R on the map as a feature.

[0192] In this embodiment, the control unit 17 uses the identified feature as the second word. That is, in this embodiment, the control unit 17 uses "National Highway No. ▽" as the second word.

[0193] [Generating search terms] The control unit 17 combines the first word, "chimney," and the second word, "National Highway No. ▽," with a space in between, to generate the search term "chimney National Highway No. ▽," which is the first search term. Then, the control unit 17 registers the generated search term in association with the utterance position.

[0194] The second word generation process performed in the control unit 17 of the in-vehicle device 10 in this embodiment will be described below. Figure 19 is a flowchart of the second word generation subroutine RT6 performed in the control unit 17 of the in-vehicle device 10. Steps S301 to S303, which perform the same processing as in Embodiment 2, will not be explained.

[0195] In step S303, the control unit 17 forms an area R on the map and then identifies features within the area R (step S601). In other words, in this embodiment, the control unit 17 identifies "National Highway No. ▽" within the area R as a feature.

[0196] The control unit 17 generates a second word based on the feature identified in step S601 (step S602). In this embodiment, the control unit 17 uses "National Highway No. ▽" as the second word. After step S602, the control unit 17 terminates the second word generation subroutine RT6.

[0197] According to this embodiment, similar to Embodiment 1, the generated search words are registered in a manner that associates them with the utterance location on the map, so that users such as passengers in vehicle M can perform searches on the search words at any time.

[0198] Furthermore, according to this embodiment, the control unit 17 identifies a feature located within a predetermined range from the utterance position, uses that feature as a second word, and generates a search word using that second word. Therefore, according to this embodiment, for example, as shown in Embodiment 4, if there are articles on the internet that introduce facilities in relation to transportation infrastructure, such articles can be searched efficiently.

[0199] In this embodiment, the control unit 17 identifies features located within a range R from the utterance position, but the number of features may be multiple. For example, the control unit 17 may identify the aforementioned "National Highway No. ▽" and the names of other road routes that intersect with "National Highway No. ▽", and use both as the second word. In this case, as explained in Embodiment 2, the search word may be generated using OR search + AND search.

[0200] In Examples 4 and 5, when the control unit 17 identifies a road route or railway route as a feature, it may generate a second word using the common name, nickname, or alternative name of the road route or railway route. For example, the control unit 17 may use the common name "Kan-nana" for "Kanjo-Nanago-sen" (Loop Route 7) or the nickname "Tobu Skytree Line" for "Tobu Isesaki Line" to generate a second word.

[0201] Furthermore, the control unit 17 may identify a road nickname based on the place name indicated by the location of the passenger's utterance in vehicle M and the name of the national highway on which vehicle M is traveling at that time. For example, if a utterance is made in Gaien while vehicle M is traveling on National Highway 246, the control unit 17 will identify "Aoyama-dori". Also, if a utterance is made in Sangenjaya while vehicle M is traveling on National Highway 246, the control unit 17 will identify "Tamagawa-dori". The control unit 17 may also use these identified road nicknames as a second word.

[0202] In Example 5, the control unit 17 identifies a feature within range R from the utterance position and generates a second word based on that feature. However, this can also be applied to the configuration of Example 3. That is, the control unit 17 may form the above-mentioned range R based on the attributes of what the common noun indicates and an expression indicating the direction from the utterance position toward what the proper noun indicates, and identify features that fall within that range R.

[0203] In Examples 1 to 5, the control unit 17 generated a second word by specifying a place name in Examples 1 to 3 and a geographical feature in Examples 4 and 5. However, it is also possible to generate the second word by including both the place name and the geographical feature.

[0204] For example, the control unit 17 may generate a second word by including "Shinjuku-ku △△-cho," which was identified as a place name in Example 1, and "○△ Station," which was identified as a feature in Example 4, to generate the search word "Ramen shop with a long line Shinjuku-ku △△-cho ○△ Station."

[0205] The control routines shown in Examples 1 to 5 above are merely illustrative and can be appropriately selected and modified depending on the application or usage conditions. [Explanation of symbols]

[0206] 10 Onboard equipment 12 microphones 13 GPS receivers 14 Touch panel display 15 speakers 16 System bus 17 Control Unit 18 Communications Department 19 Input section 21 Output section 22 Mass storage

Claims

1. A voice acquisition unit that acquires input audio through user utterances, A speech position acquisition unit that acquires the speech position at which the utterance was made by the user, A map data acquisition unit that acquires map data, A unit that identifies one or more place names or features based on the utterance location and the map data, The system includes a search word generation unit that generates a search word comprising a first word generated based on the input voice and a second word generated based on the one or more place names or features identified by the identification unit, The information processing device is characterized in that the identifying unit converts the input voice into a string, identifies a feature indicated by a proper noun contained in the converted string, identifies the direction from the utterance position to the location of the feature as a reference direction by referring to the map data, identifies a directional expression which is a word indicating the direction relative to the reference direction from the string, and identifies one or more place names or features based on the reference direction and the directional expression.

2. The identification unit converts the input audio into a string, identifies the feature or event indicated by the common noun contained in the converted string, and identifies an attributive modifier, attributive phrase, or attributive clause related to the common noun from the string. The information processing apparatus according to claim 1, characterized in that the search word generation unit generates the first word consisting of a feature or event indicated by the common noun and the attributive modifier, attributive phrase, or attributive clause.

3. The information processing device according to claim 2, characterized in that the identifying unit identifies one or more place names or features based on a distance range determined according to the attributes of the feature or event indicated by the common noun.

4. The information processing device according to any one of claims 1 to 3, characterized in that the aforementioned features include railway lines or road lines.

5. An information processing method performed by an information processing device, A speech acquisition step that obtains input audio input through user utterance, A speech position acquisition step to acquire the speech position at which the utterance was made by the user, Map data acquisition step to acquire map data, A selection step to identify one or more place names or features based on the utterance location and the map data, A search word generation step is provided for generating a search word that includes a first word generated based on the input voice and a second word generated based on the one or more place names or features identified by the specific step. An information processing method characterized in that, in the specified step, the input speech is converted into a string, a feature indicated by a proper noun contained in the converted string is identified, the direction from the utterance position to the location of the feature is identified as a reference direction by referring to the map data, a directional expression which is a word indicating the direction relative to the reference direction is identified from the string, and one or more place names or features are identified based on the reference direction and the directional expression.

6. A program to be executed by a computer, A speech acquisition step that obtains input audio input through user utterance, A speech position acquisition step to acquire the speech position at which the utterance was made by the user, Map data acquisition step to acquire map data, A selection step to identify one or more place names or features based on the utterance location and the map data, The computer is made to perform a search word generation step which generates a search word that includes a first word generated based on the input voice and a second word generated based on the one or more place names or features identified by the specific step. A program characterized in that, in the specified step, it converts the input speech into a string, identifies a feature indicated by a proper noun contained in the converted string, identifies a reference direction from the utterance position to the location of the feature by referring to the map data, identifies a directional expression which is a word indicating the direction relative to the reference direction from the string, and identifies one or more place names or features based on the reference direction and the directional expression.

7. A recording medium storing the program described in claim 6.

Images