JP6781961B2 - Transmission device, transmission method and transmission program - Google Patents

Description

The present invention relates to a transmitting device, a transmitting method and a transmitting program.

As a conventional information presenting device, there is one described in Japanese Patent Application Laid-Open No. 2012-256183 (Patent Document 1). The conventional information presenting device estimates the strength of the desire that the user currently has and the situation that the user is currently in, and queries the database that describes the combination of these, so that the user has the information. Present items that can satisfy both the desires of the user and the situation in which the user is placed.

Japanese Unexamined Patent Publication No. 2012-256183

However, since the conventional information presenting device only estimates the strength of the user's desire and the situation in which the user is placed, it is not always possible to present information according to the user's needs.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide information according to the needs of a user.

The search program according to one aspect of the present invention is a search program that causes a computer connected to a transmission device to function as a search device, and the transmission device is a voice collection unit that collects voice and generates voice data indicating the voice. And, from the voice data, a voice keyword extraction unit that extracts a voice keyword that is a keyword included in the voice, and from the voice data, extracts a voice feature including at least one of voice strength, speed, and intonation. A voice feature extraction unit, an emotion keyword selection unit that selects an emotion keyword corresponding to the extracted voice characteristics from a group of emotion keywords stored in advance based on the voice characteristics, and an extracted voice keyword and selected It includes a transmitter that sends emotional keywords to a computer and a dictionary database in which a plurality of words are stored in advance. The voice keyword extraction unit matches any of the words contained in the voice with any of the words stored in the dictionary database. In that case, the word is extracted as a voice keyword, the computer executes a search process based on the voice keyword and the emotion keyword, and the search program gives the computer an instruction to update the keyword database for storing the voice keyword and the dictionary database. Is stored in the keyword database, a receiving unit that receives the data from the transmitting device, a guessing unit that infers the related words of the voice keywords stored in the keyword database in response to the update instruction received through the receiving unit, and the keyword database. A dictionary database creation unit that creates an updated dictionary database based on voice keywords and inferred related words, and a transmission unit that transmits the updated dictionary database to a transmission device are realized.

According to the present invention, it is possible to provide information according to the needs of the user.

It is a figure which shows the structure of the search system 100 including the transmission device 110 which concerns on one Embodiment of this invention. It is a figure which shows an example of the functional block included in the transmission device 110. It is a figure which shows the emotion ring model of Russell and the keyword contained in it. It is a figure which shows an example of the functional block included in the search apparatus 120. It is a figure which shows an example of the functional block included in the output device 130. It is a flowchart which shows an example of the process in the search system 100. It is a flowchart which shows the dictionary database update process.

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. FIG. 1 is a diagram showing a configuration of a search system 100 including a transmission device 110 according to an embodiment of the present invention. The search system 100 includes a transmission device 110, a search device 120, and an output device 130. The search system 100 according to the present embodiment performs a predetermined search based on the keyword extracted from the voice collected by the transmission device 110 and the keyword corresponding to the emotion or atmosphere extracted from the voice, and the search result is obtained. It is a system that outputs.

The transmission device 110 is a computer that transmits a predetermined keyword to the search device 120 based on the voice collected at the place where the transmission device 110 is placed. The transmission device 110 may be a computer installed in a facility or store, or may be a computer (smartphone, tablet terminal, or the like) owned by a user who visits the facility or store. The transmitting device includes a processor, memory and a communication interface. The transmission device 110 can communicate with the search device 120 via, for example, a mobile phone communication network or the Internet.

The search device 120 is a computer (server) that provides information to the user of the output device 130 based on the keyword received from the transmission device 110. The search device 120 includes a processor, memory, a database, and a communication interface. The search device 120 can communicate with the transmission device 110 and the output device 130, for example, via the Internet.

The output device 130 is a computer that outputs search results based on the data (display data) provided by the search device 120. The output device 130 outputs numerical values, characters, video (images), audio, etc. as search results to a display, a speaker, or the like. The output device 130 is, for example, a smartphone, a tablet terminal, a personal computer, or the like. The output device 130 includes a processor, a memory, and a communication interface. The output device 130 can communicate with the search device 120 via, for example, a mobile phone communication network or the Internet.

The output device 130 may output a predetermined operation based on the search result. The output device 130 may, for example, perform predetermined communication or control a motor, an actuator, a sensor, or the like based on the search result.

FIG. 2 is a diagram showing an example of a functional block included in the transmission device 110. The transmission device 110 includes a voice collection unit 200, a voice keyword extraction unit 210, a dictionary database 220, a voice feature extraction unit 230, an emotion keyword selection unit 240, an emotion database 250, a transmission unit 260, and a dictionary database update. It is provided with an instruction creation unit 270.

A transmission program is stored in the memory of the transmission device 110, and each function of the transmission device 110 (voice collection unit 200, voice keyword extraction unit 210, etc.) is stored in cooperation with the hardware resources of the transmission device 110 and the transmission program. The dictionary database 220, the voice feature extraction unit 230, the emotion keyword selection unit 240, the emotion database 250, the transmission unit 260, and the dictionary database update instruction creation unit 270) are realized. The transmission program is read from a computer-readable recording medium into the memory of the transmission device 110 and executed by the processor of the transmission device 110.

The voice collecting unit 200 collects the voice emitted around the transmitting device 110 and generates voice data indicating the voice. Specifically, the voice collecting unit 200 converts the voice collected by the microphone or the like into an electric signal, and generates the voice data obtained by converting the information indicated by the electric signal into digital data. Further, the voice collecting unit 200 may have a storage unit that temporarily stores the generated voice data. The storage unit is an existing storage unit that can store magnetically, electrically or optically, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), a memory card, an optical disk, or a RAM (Random Access Memory). A device or storage medium. The voice collecting unit 200 may be the built-in microphone of the transmitting device 110, or may be an external microphone (external wired microphone or wireless microphone). For example, a tablet terminal to which an external microphone is connected functions as a transmitting device 110. The transmission device 110 may include a plurality of voice collecting units 200. The transmitting device 110 may detect the relative positions of the plurality of voice collecting units 200.

The voice keyword extraction unit 210 extracts a keyword (hereinafter, also referred to as “voice keyword”) included in the voice from the voice data generated by the voice collection unit 200. Specifically, the voice keyword extraction unit 210 first analyzes the voice data and converts the voice into text data. Then, the voice keyword extraction unit 210 compares each word included in the text data with a word stored in advance in the dictionary database 220. Then, when each word included in the text data matches a word stored in advance in the dictionary database 220, the voice keyword extraction unit 210 extracts the word as a voice keyword.

The voice feature extraction unit 230 extracts the voice feature from the voice data generated by the voice collection unit 200. In the present embodiment, the voice feature extraction unit 230 extracts at least one of the voice strength, speed and intonation, and the strength, speed or intonation, or a combination of two or more thereof. Extract as a feature of the voice.

The voice feature extraction unit 230 extracts the strength of the voice based on the magnitude of the amplitude of the voice signal indicated by the voice data. The voice feature extraction unit 230 extracts, for example, the average voice strength in a predetermined unit period as the voice strength. Further, the voice feature extraction unit 230 may extract the average of the voice intensities including a predetermined number of characters, words, or sentences as the voice strength.

Further, the voice feature extraction unit 230 extracts the speed of the voice based on the number of sounds of the words included in the voice. The voice feature extraction unit 230 extracts, for example, the number of sounds included in a predetermined unit period as the speed of the voice. In addition, the voice feature extraction unit 230 may extract the number of characters, the number of words, or the number of sentences of the voice in a predetermined unit period as the speed of the voice.

In addition, the voice feature extraction unit 230 extracts the intonation of the voice based on the change in the strength of the voice. For example, the voice feature extraction unit 230 divides a voice into a plurality of units, and extracts a change in strength within each unit and / or a change in strength between units as intonation of the voice. A voice unit is, for example, a word or a clause that constitutes a sentence included in the voice.

The emotion keyword selection unit 240 converts the speaker's emotions and the atmosphere of the place indicated by the characteristics into keywords based on the voice characteristics extracted by the voice feature extraction unit 230. The emotion keyword selection unit 240 selects, for example, an emotion keyword corresponding to a voice feature extracted by the voice feature extraction unit 230 from a group of emotion keywords stored in advance in the emotion database 250. In addition, the emotion database 250 stores a large number of keywords related to emotions and moods in association with voice features. Specifically, the emotion database 250 associates each value (or a predetermined range in which the value can be taken) of the voice intensity, speed, and intonation, and a pattern of the combination of the value or range with each keyword. And store. Then, the emotion keyword selection unit 240 sets the voice features extracted by the voice feature extraction unit 230, that is, the respective values (or a predetermined range in which the values can be taken) of the voice strength, speed, and intonation, and the values. Alternatively, a predetermined keyword is selected from a large number of keywords stored in the emotion database 250 based on the pattern of the combination of ranges. The keyword stored in the emotion database 250 may be a keyword included in the so-called Russell's emotion ring model as shown in FIG.

The emotion keyword selection unit 240 may select an emotion keyword based on the biological information of the speaker who emitted the collected voice, in addition to the voice features extracted by the voice feature extraction unit 230. The emotion keyword selection unit 240 estimates the emotion of the speaker based on biological information such as the speaker's body temperature, heartbeat, pulse, brain wave, and skin conductance level, and selects a keyword corresponding to the estimated emotion. You may. The biological information of the speaker may be acquired from a sensor connected to the transmitting device 110, or may be acquired from an external device of the transmitting device 110, for example, through a network.

The voice feature extraction unit 230 may extract an index relating to the degree of affirmation or denial of speech from the waveform of voice data. For example, the voice feature extraction unit 230 can use Empath (registered trademark) to extract an index relating to the degree of affirmation or denial of speech from the waveform of voice data. Empath analyzes the physical characteristics of the waveform of the voice data, and calculates the judgment value in 50 steps for each item of normality, anger, joy, sadness, and energy level. The voice feature extraction unit 230 can extract an index relating to the degree of affirmation or denial of the statement based on the result calculated in this way. In this way, the technique of calculating the judgment value regarding emotions and moods from the waveform of the voice data is called the voice mood analysis technique. The items determined using the voice mood analysis technique are not limited to normality, anger, joy, sadness and energy, but may include all items related to emotions. The emotion database 250 may store an index relating to the degree of affirmation or denial of a statement in association with each emotion keyword. The emotion keyword selection unit 240 may select an emotion keyword from the emotion database 250 that corresponds to an index related to the degree of affirmation or denial of speech extracted by the voice feature extraction unit 230.

The transmission unit 260 transmits the voice keyword extracted by the voice keyword extraction unit 210 and the emotion keyword selected by the emotion keyword selection unit 240 to the search device 120. In the transmission unit 260, for example, a predetermined period has passed since the voice collection unit 200 collected the voice, the strength of the recorded voice extracted by the voice feature extraction unit 230 exceeded the predetermined value, and the voice keyword. In response to the extraction of a predetermined word by the extraction unit 210 or the like, the voice keyword extracted by the voice keyword extraction unit 210 and the emotion keyword selected by the emotion keyword selection unit 240 are transmitted to the search device 120. The predetermined word is, for example, "search" or the like.

The transmission unit 260 may further transmit environmental information or biological information to the search device 120 in addition to the voice keyword and the emotion keyword. The environmental information is information about a place or space where the transmitting device 110 is placed, such as a position, temperature, humidity, illuminance, shaking, and the like. Environmental information is acquired by measuring devices such as GPS, temperature sensor, humidity sensor, illuminance sensor, acceleration sensor, and infrared sensor. The transmission unit 260 may acquire environmental information and biological information directly from the measuring device, or indirectly through a network or the like.

The dictionary database update instruction creation unit 270 creates an instruction (dictionary database update instruction) for updating the dictionary database 220. The dictionary database update instruction is created at a predetermined opportunity. Here, the predetermined trigger may be when the hit rate of the dictionary database 220 falls below a predetermined threshold value, or may be a predetermined fixed interval. The hit rate means the ratio of keywords that have been determined to be included in the voice collected by the voice collecting unit 200 among all the keywords registered in the dictionary database 220. The transmission unit 260 transmits the dictionary database update instruction created by the dictionary database update instruction creation unit 270 to the search device 120.

FIG. 4 is a diagram showing an example of a functional block included in the search device 120. The search device 120 includes a receiving unit 300, a search executing unit 310, a database 320, a transmitting unit 330, a keyword database 340, a guessing unit 350, and a dictionary database creating unit 360. The search device 120 is a search engine that searches for information stored in the database 320 based on a predetermined keyword and transmits the search result to the output device 130. The search device 120 may receive the voice keyword transmitted from the transmission device 110 and store the voice keyword to perform a process of estimating a topic from the stored voice keyword. When the voice keywords are, for example, "handle", "brake", "tire", etc., it can be inferred that the topic of "car" is being made. The inferred topic is provided to, for example, an advertiser. The search device 120 also performs a process of updating the dictionary database 220 in response to a dictionary database update instruction from the transmission device 110.

A search program for causing the search device 120 to execute a search process based on a voice keyword and an emotional keyword is stored in the memory of the search device 120, and the search device cooperates with the hardware resources of the search device 120 and the search program. Each of the 120 functions (reception unit 300, search execution unit 310, database 320, transmission unit 330, keyword database 340, guessing unit 350, and database creation unit 360) is realized. The search program is read into the memory of the search device 120 from a computer-readable recording medium and executed by the processor of the search device 120.

The receiving unit 300 receives the voice keyword and the emotion keyword transmitted by the transmitting device 110. Further, the receiving unit 300 may further receive environmental information and biological information in addition to the voice keywords and emotional keywords transmitted by the transmitting device 110. The receiving unit 300 receives the dictionary database update instruction transmitted by the transmitting device 110.

The search execution unit 310 searches the database 320 based on the voice keyword and emotion keyword received by the reception unit 300 and / or the environmental information and / or the biological information. In the present embodiment, the voice keyword, the emotion keyword, the environmental information, and the biological information are all text data, and the search execution unit 310 uses, for example, the text data used for the search among the information stored in the database 320. Information including any of the above is extracted as a search result. The information is, for example, text data included in a website.

The transmission unit 330 transmits the search result extracted by the search execution unit 310 to the output device 130 via the network. For example, the transmission unit 330 transmits the URL of the website extracted by the search execution unit 310, text data, image data, and the like included in the website to the output device 130 as a search result.

The keyword database 340 stores voice keywords received through the receiving unit 300. The guessing unit 350 infers the related words of the voice keywords stored in the keyword database 340 in response to the dictionary database update instruction received through the receiving unit 300. A related word of a voice keyword means a word that is frequently used at the same time as a voice keyword, and such a word is also called a “co-occurrence word”. Co-occurrence words do not necessarily mean synonymous words. The guessing unit 350 is, for example, a co-occurrence word search tool that applies artificial intelligence. As such a co-occurrence word search tool, for example, a tool that performs morphological analysis on the content of a web page such as Google or Yahoo that ranks higher in the search results and presents words and phrases that frequently appear in the same document as co-occurrence words is known. ing. Morphological analysis is a sequence of morphological elements (the smallest unit that has meaning in a language) based on information such as the grammar of the target language and the part of speech of a word called a dictionary, from text data in a natural language without notes of grammatical information. It means the work of decomposing into and determining the part of speech of each morphological element.

The dictionary database creation unit 360 updates the dictionary database 220 in response to the dictionary database update instruction received through the reception unit 300. The dictionary database creation unit 360 creates an updated dictionary database 220 based on the voice keywords stored in the keyword database 340 and the related words inferred by the guessing unit 350. The updated dictionary database 220 registers voice keywords that have been extracted from the voice collected by the voice collection unit 200 and related words as keywords. Among the keywords registered in the dictionary database 220 before the update, the voice keywords that have not been extracted from the voice collected by the voice collection unit 200 are deleted from the dictionary database 220 after the update. By repeating such an update process, the hit rate of the dictionary database 220 can be increased.

The transmission unit 330 transmits the updated dictionary database 220 to the transmission device 110. The transmitting device 110 that has received the updated dictionary database 220 replaces the dictionary database 220 before the update with the dictionary database 220 after the update.

FIG. 5 is a diagram showing an example of a functional block included in the output device 130. The output device 130 includes a search result receiving unit 400 and a search result output unit 410. In the output device 130, the search result receiving unit 400 receives the search result transmitted by the search result transmitting unit 330, and the search result output unit 410 receives the received search result in numerical values, characters, video (image), voice, etc. Is output through a display, a speaker, or the like.

An output program is stored in the memory of the output device 130, and each function of the output device 130 (search result receiving unit 400 and search result output unit) is performed in cooperation with the hardware resources of the output device 130 and the output program. 410) is realized. The output program is read from a computer-readable recording medium into the memory of the output device 130 and executed by the processor of the output device 130.

FIG. 6 is a flowchart showing an example of processing in the search system 100.

First, the voice collecting unit 200 collects the voice emitted around the transmitting device 110 and generates voice data indicating the voice (S600). The voice collecting unit 200 generates this voice data when, for example, the voice collecting unit 200 collects the voice "It is hot today" from one or a plurality of speakers. Next, the voice keyword extraction unit 210 extracts voice keywords from the voice data (S601). The voice keyword extraction unit 210 extracts the word "hot" as a voice keyword from the voice data such as "It's hot today." Next, the voice feature extraction unit 230 extracts the voice feature from the voice data generated by the voice collection unit 200 (S602). The voice feature extraction unit 230 extracts, for example, the strength, speed, and intonation of the voice in the voice "It's hot today."

Next, the emotion keyword selection unit 240 selects an emotion keyword indicating the speaker's emotion and the atmosphere of the place based on the voice characteristics extracted by the voice feature extraction unit 230 (S603). The emotion keyword selection unit 240 is stored in the emotion database 250 as an emotion keyword corresponding to a pattern that combines the strength, speed, and intonation value of the voice in the voice "It's hot today." Select the "irritated" that was. Next, the transmission unit 260 transmits the voice keyword extracted by the voice keyword extraction unit 210 and the emotion keyword selected by the emotion keyword selection unit 240 to the search device 120 (S604). The transmission unit 260 transmits "hot" and "irritated" to the search device 120 as voice keywords and emotion keywords, respectively. In addition, the transmission unit 260 provides location information indicating "outside Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture" where the transmitter 110 is located, and "36 ° C" which is the temperature at the location. , Is transmitted to the search device 120 as environmental information.

Next, the keyword receiving unit 300 receives the voice keyword, the emotion keyword, and the environmental information transmitted by the transmitting device 110 (S605). The keyword receiving unit 300 uses, for example, "hot" and "irritated" as voice keywords and emotional keywords, respectively, and position information and temperature "36" indicating "outside Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture". ℃ ”is received as environmental information. Next, the search execution unit 310 searches the database 320 based on the voice keyword, the emotion keyword, and the environmental information received by the keyword reception unit 300 (S606). In the search execution unit 310, for example, based on these keywords and environmental information, the speaker who uttered the voice "It's hot today." Is "outside Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture." In, it is judged that the word "hot" is emitted together with the feeling of "irritability" in the temperature "36 ° C.". Then, the search execution unit 310 searches for "a store that offers cold drinks within a radius of 100 m of" Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture ". Next, the search result transmission unit 330 transmits the search result extracted by the search execution unit 310 to the output device 130 via the network (S607). The search result transmission unit 330 outputs, for example, a list of websites of stores corresponding to "a store that provides cold drinks within a radius of 100 m of" Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture "as a search result. It is transmitted to the device 130.

Next, the search result receiving unit 400 receives the search result transmitted by the search result transmitting unit 330 (S608). Then, the search result output unit 410 outputs the received search result through a display, a speaker, or the like (S609). The search result output unit 410 displays, for example, a list of websites of stores corresponding to "a store that provides cold drinks within a radius of 100 m of" Nagaokakyo Station in 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture "". Display on the provided display. In this example, for example, the location information indicates "a building at 1-10-1, Higashikotari, Nagaokakyo City, Kyoto Prefecture" instead of "outside Nagaokakyo Station, located at 1-chome, Kotari, Nagaokakyo City, Kyoto Prefecture", and the temperature is "36". When indicating "28 ° C" instead of "° C", the search execution unit 310 tells the speaker who uttered the voice "It's hot today" to use the word "hot" in the building as the temperature "28 ° C". It may be judged that it is emitted with the feeling of "irritability". In this case, the search execution unit 310 may search for "the temperature can be controlled in the building" and output that the air conditioning is controlled as a search result. In this case, the search result output unit 410 outputs, for example, changing the set temperature of the air conditioner to "25 ° C.".

Next, the flow of the dictionary database update process will be described with reference to FIG. 7.
The transmission device 110 creates a dictionary database update instruction (S700) at a predetermined opportunity, and transmits this to the search device 120 (S701). When the search device 120 receives the dictionary database update instruction (S702), it acquires the voice keyword from the keyword database 340 (S703) and infers the related word of the voice keyword (S704). Next, the search device 120 creates an updated dictionary database 220 based on the voice keyword and its related words (S705), and transmits this to the transmission device 110 (S706). When the transmitting device 110 receives the updated dictionary database 220 (S707), the transmitting device 110 updates the dictionary database 220 by replacing the old dictionary database 220 with the updated dictionary database 220 (S708).

The exemplary embodiments of the present invention have been described above. According to the present embodiment, voice is collected to generate voice data indicating the voice, voice keywords that are keywords included in the voice are extracted from the voice data, and voice strength, speed, and voice data are extracted from the voice data. Voice features including at least one of intonations are extracted, and based on the voice features, emotion keywords corresponding to the extracted voice features are selected from a group of emotion keywords stored in advance, and the extracted voice. Send keywords and selected emotion keywords. As a result, it is possible to search based on both words and emotions extracted from the voice, so that it is possible to provide information according to the needs of the speaker who uttered the voice.

Further, in the present embodiment, the keywords included in Russell's emotional ring model may be stored as a keyword group. As a result, the emotions of the speaker can be extracted more accurately.

Further, in the present embodiment, when a word included in the voice matches any of the words stored in the dictionary database, the word may be extracted as a voice keyword. As a result, the load of the algorithm for extracting the voice keyword can be reduced.

Further, in the present embodiment, a predetermined period has passed since the voice collecting unit collected the voice, the strength of the voice extracted by the voice feature extraction unit exceeded the predetermined value, or the voice keyword extraction. In response to the department extracting a predetermined word, a voice keyword and a selected emotion keyword may be transmitted. As a result, the accuracy of the search can be further improved.

Further, in the present embodiment, a search result by a predetermined search may be output. As a result, the search result can be transmitted to the speaker or reflected in the environment in which the speaker is placed.

The dictionary database 220 is updated by updating the dictionary database 220 based on the voice keywords extracted from the voice collected by the voice collection unit 200 and their related words as matching the voice keywords registered in the dictionary database 220. You can increase the hit rate of. As a result, the dictionary database 220 can be optimized. By optimizing the dictionary database 220, it is appropriately determined whether or not the product or service is actually popular based on the voice collected by the voice collecting unit 200 (for example, the voice actually spoken by the customer). be able to. Further, since the voice collected by the voice collecting unit 200 is the voice actually collected at the place where the transmitting device 110 is placed, the product or service is based on the evaluation in the real world, not the evaluation on the Internet. Can be determined whether or not is actually in fashion. Further, since unnecessary keywords are deleted from the dictionary database 220 by repeating the update of the dictionary database 220, the storage capacity of the dictionary database 220 may be small. This makes it possible to reduce the size and power consumption of the transmission device 110.

It should be noted that each of the embodiments described above is for facilitating the understanding of the present invention, and is not for limiting the interpretation of the present invention. The present invention can be modified / improved without departing from the spirit thereof, and the present invention also includes an equivalent thereof. That is, those skilled in the art with appropriate design changes to each embodiment are also included in the scope of the present invention as long as they have the features of the present invention. For example, each element included in each embodiment and its arrangement, material, condition, shape, size, and the like are not limited to those exemplified, and can be appropriately changed. In addition, each embodiment is an example, and it goes without saying that the configurations shown in different embodiments can be partially replaced or combined, and these are also included in the scope of the present invention as long as the features of the present invention are included. ..

100 ... Search system, 110 ... Transmission device, 120 ... Search device, 130 ... Output device, 200 ... Voice collection unit, 210 ... Voice keyword extraction unit, 220 ... Dictionary database, 230 ... Voice feature extraction unit, 240 ... Emotion keyword selection Unit, 250 ... Emotion database, 260 ... Transmission unit, 300 ... Keyword receiving unit, 310 ... Search execution unit, 320 ... Database, 330 ... Search result transmitting unit, 400 ... Search result receiving unit, 410 ... Search result output unit

Claims (12)

A search program that allows the computer connected to the transmitter to function as a search device.
The transmitter is
A voice collection unit that collects voice and generates voice data indicating the voice,
A voice keyword extraction unit that extracts voice keywords that are keywords included in the voice from the voice data,
A voice feature extraction unit that extracts the voice features including at least one of the voice strength, speed, and intonation from the voice data.
An emotion keyword selection unit that selects an emotion keyword corresponding to the extracted voice characteristics from a group of emotion keywords stored in advance based on the voice characteristics.
A transmitter that transmits the extracted voice keyword and the selected emotion keyword to the computer .
A dictionary database with multiple words stored in advance and
With
When the word included in the voice matches any of the words stored in the dictionary database, the voice keyword extraction unit extracts the word as the voice keyword.
The computer executes a search process based on the voice keyword and the emotion keyword, and performs search processing.
The search program is applied to the computer.
A keyword database that stores the voice keywords and
A receiving unit that receives an update instruction of the dictionary database from the transmitting device, and
In response to the update instruction received through the receiving unit, the guessing unit that infers the related words of the voice keywords stored in the keyword database, and
A dictionary database creation unit that creates an updated dictionary database based on the voice keywords stored in the keyword database and the inferred related words.
A transmitter that transmits the updated dictionary database to the transmitter,
A search program that realizes. The search program according to claim 1.
The transmission device is a search program further including an emotion database that stores keywords included in Russell's emotion ring model as the keyword group. The search program according to claim 1 or 2.
The transmitting unit is a search program that transmits the voice keyword and the selected emotion keyword to the computer in response to the elapse of a predetermined period after the voice collecting unit collects the voice. The search program according to any one of claims 1 to 3.
The voice feature extraction unit extracts at least the strength of the voice as a feature of the voice.
The transmission unit is a search program that transmits the voice keyword and the selected emotion keyword to the computer in response to the voice intensity extracted by the voice feature extraction unit exceeding a predetermined value . .. The search program according to any one of claims 1 to 4.
The transmission unit is a search program that transmits the voice keyword and the selected emotion keyword to the computer in response to the voice keyword extraction unit extracting a predetermined word. The search program according to any one of claims 1 to 5.
The computer is a search program connected to an output device that outputs an execution result of the search process. A search device that connects to a transmitter
The transmitter is
A voice collection unit that collects voice and generates voice data indicating the voice,
A voice keyword extraction unit that extracts voice keywords that are keywords included in the voice from the voice data,
A voice feature extraction unit that extracts the voice features including at least one of the voice strength, speed, and intonation from the voice data.
An emotion keyword selection unit that selects an emotion keyword corresponding to the extracted voice characteristics from a group of emotion keywords stored in advance based on the voice characteristics.
A transmission unit that transmits the extracted voice keyword and the selected emotion keyword to the search device, and
A dictionary database with multiple words stored in advance and
With
When the word included in the voice matches any of the words stored in the dictionary database, the voice keyword extraction unit extracts the word as the voice keyword.
The search device executes a search process based on the voice keyword and the emotion keyword, and performs search processing.
The search device is
A keyword database that stores the voice keywords and
A receiving unit that receives an update instruction of the dictionary database from the transmitting device, and
In response to the update instruction received through the receiving unit, the guessing unit that infers the related words of the voice keywords stored in the keyword database, and
A dictionary database creation unit that creates an updated dictionary database based on the voice keywords stored in the keyword database and the inferred related words.
A transmitter that transmits the updated dictionary database to the transmitter,
A search device. A search method performed on the computer connected to the transmitter.
The transmitter is
Collecting voice and generating voice data indicating the voice,
Extracting a voice keyword, which is a keyword included in the voice, from the voice data,
Extracting the characteristics of the voice, including at least one of the strength, speed and intonation of the voice, from the voice data.
To select an emotional keyword corresponding to the extracted voice characteristic from a group of emotional keywords stored in advance based on the voice characteristic.
Sending the extracted voice keyword and the selected emotion keyword to the computer, and
To realize a dictionary database in which multiple words are stored in advance,
And run
Extracting the voice keyword includes extracting the word as the voice keyword when the word contained in the voice matches any of the words stored in the dictionary database.
The computer executes a search process based on the voice keyword and the emotion keyword, and performs search processing.
The computer
A keyword database for storing the voice keywords has been realized.
Upon receiving the update instruction of the dictionary database from the transmitter,
In response to the received update instruction, the related word of the voice keyword stored in the keyword database is inferred.
An updated dictionary database is created based on the voice keywords stored in the keyword database and the inferred related words.
A search method for transmitting the updated dictionary database to the transmitting device. The search method according to claim 8.
A search method in which the transmitting device transmits the position information of the transmitting device to the computer . The search program according to any one of claims 1 to 6, wherein the guessing unit guesses a related word of the voice keyword by applying artificial intelligence. The search device according to claim 7, wherein the guessing unit guesses a related word of the voice keyword by applying artificial intelligence. The search method according to claim 8 or 9, wherein guessing a related word of the voice keyword includes guessing a related word of the voice keyword by applying artificial intelligence.

Images