JP6597156B2 - Information generation system - Google Patents

Description

  The present invention relates to a technique for generating information from guidance voice.

  Techniques have been conventionally proposed that provide various types of information to users according to the results of speech recognition for uttered speech. For example, Patent Literature 1 discloses a configuration in which voice recognition is performed on a user's uttered voice and a map display area, a display scale, and the like are instructed around a target point specified as a result of voice recognition. Yes.

Japanese Patent Laid-Open No. 03-175478

  By the way, for example, if the guidance voice broadcast in a commercial facility such as a shopping mall can be recognized and related information such as a character string of the recognition result and its translation can be provided to the terminal device of the facility visitor, the guidance voice can be heard. Useful for people who have difficulty hearing or who cannot understand the language of the guidance voice. However, it is actually difficult to realize complete speech recognition. For example, erroneous recognition may occur due to utterance characteristics peculiar to individual speakers or background noise at the time of sound collection. . In view of the above circumstances, an object of the present invention is to provide a user with relevant information appropriate to a guidance voice.

  In order to solve the above problems, an information generation system according to the present invention is a system that generates distribution information transmitted to a terminal device so that the terminal device presents related information related to the guidance voice to the user. In addition, for each of a plurality of recognized character strings analyzed by voice recognition for the guidance voice, a character string specification that identifies the registered character string that is similar to the recognized character string among a plurality of registered character strings that represent different pronunciation contents And the terminal device provides related information corresponding to the correspondence information corresponding to the combination of the plurality of registered character strings specified by the character string specifying unit among the plurality of pieces of correspondence information designating different combinations of the plurality of registered character strings. An information generation unit that generates distribution information to be presented to the user. In the above configuration, a registered character string similar to a recognized character string analyzed by voice recognition for the guidance voice is specified from among a plurality of registered character strings. Therefore, it does not include the influence of misrecognition of voice recognition, compared with a configuration that generates distribution information for presenting related information according to a plurality of recognized character strings analyzed by voice recognition to the guidance voice. It is possible to generate distribution information for presenting related information to the user. In addition, since distribution information for the terminal device to present to the user related information corresponding to correspondence information specifying different combinations of a plurality of registered character strings is generated, a plurality of registrations specified by the character string specifying unit Compared to a configuration that generates distribution information for presenting relevant information according to a character string to the user, for example, when the guide mispronounces the guidance voice in units of sentences (for example, to pronounce the necessary sentences) For presenting appropriate related information according to the corresponding information corresponding to the combination of multiple registered character strings specified from the recognition result of the guidance voice, even when forgetting or when an unnecessary sentence is pronounced) Distribution information can be generated.

  In a preferred aspect of the present invention, the information generation unit generates a guide sentence of correspondence information corresponding to a combination of a plurality of registered character strings specified by the character string specifying part among a plurality of guide sentences prepared for each piece of correspondence information. Distribution information for the terminal device to present the related information to be shown to the user is generated. In the above configuration, distribution information is generated for the terminal device to present related information indicating the guidance text of the corresponding information corresponding to the combination of the plurality of registered character strings specified by the character string specifying unit to the user. Therefore, it is possible to generate distribution information for presenting the user with related information indicating a guide sentence prepared independently of the combination of registered character strings designated by the correspondence information.

  In a preferred aspect of the present invention, the information generation unit is configured such that the terminal device displays related information indicating a plurality of registered character strings specified by correspondence information corresponding to a combination of a plurality of registered character strings specified by the character string specifying unit. To generate distribution information for presentation. With the above configuration, distribution information for the terminal device to present related information indicating a plurality of registered character strings designated by the correspondence information to the user is generated. Therefore, in comparison with the configuration for generating the distribution information for presenting the related information indicating the guidance text of the corresponding information corresponding to the combination of the plurality of registered character strings identified by the character string identification section, the guidance text is There is an advantage that information for uniquely specifying is unnecessary.

  In a preferred aspect of the present invention, the correspondence information specifies a combination of a plurality of registered character strings and an order of the plurality of registered character strings related to the combination, and the information generation unit is arranged in the order specified by the correspondence information The terminal device generates distribution information for presenting the relevant information indicating the plurality of registered character strings to the user. In the above configuration, distribution information for the terminal device to present related information indicating a plurality of registered character strings arranged in the order designated by the correspondence information to the user is generated. Therefore, regardless of the order of pronunciation of the plurality of sentences constituting the guidance voice, the distribution information for presenting the related information indicating the plurality of registered character strings in which the correspondence information is arranged in the order specified in advance is generated. Is possible.

  The suitable aspect of this invention comprises the sound emission part which emits a guidance audio | voice and the sound which shows delivery information. In the above configuration, the sound emitting unit that emits the guidance voice is used for sound emission of distribution information (that is, acoustic communication using sound that is air vibration as a transmission medium). Therefore, it is possible to simplify the configuration of the information generation system as compared to a configuration in which a sound emitting device separate from the sound emitting device used for sounding the guidance voice is used for sound emission indicating the distribution information. It is.

1 is a schematic diagram of an information generation system according to a first embodiment. It is a block diagram of a voice guidance system and a management apparatus. It is explanatory drawing of the operation | movement which divides | segments the analysis result of a guidance voice into a some recognition character string. It is a schematic diagram of a character string table. It is a schematic diagram of a guidance table. It is a block diagram of a terminal device. It is explanatory drawing of operation | movement of an information generation system. It is a schematic diagram of the guidance table which concerns on 2nd Embodiment.

<First Embodiment>
FIG. 1 is a configuration diagram of an information generation system 1 according to the first embodiment of the present invention. The information generation system 1 of the first embodiment is a computer system for generating information to be provided to a user U in a commercial facility 300 such as a shopping mall, and includes a voice guidance system 100 and a management device 10. . The voice guidance system 100 is installed in, for example, a commercial facility 300 and communicates with the management apparatus 10 via a communication network 200 including the Internet. The management device 10 is a server (for example, a web server) connected to the communication network 200, for example. The user U carries the terminal device 30 and visits the commercial facility 300. The terminal device 30 is a portable information processing device such as a mobile phone or a smartphone. In practice, a plurality of users U in the commercial facility 300 can use the service of the information generation system 1, but in the following description, attention is focused on one terminal device 30 for convenience.

  The guide OP, such as an employee of the commercial facility 300, produces various voices (hereinafter referred to as “guide voice”) V that guide the user U who is a visitor. In the first embodiment, information for visitors of the commercial facility 300 composed of a plurality of sentences, for example, “Thank you for coming today. It will be closing time soon. Please be careful. Is pronounced as the guidance voice V. In the first embodiment, it is assumed that the guider OP selectively pronounces any of a plurality of sentences recorded in advance in an announcement book or the like to be referred to when the guidance voice V is sounded as the guidance voice V. Each of the plurality of sentences is configured by combining one sentence-unit character string (hereinafter referred to as “registered character string”) representing different utterance contents.

  The information generation system 1 generates information (hereinafter referred to as “distribution information”) Q for presenting information (hereinafter referred to as “related information”) Z related to the guidance voice V pronounced by the guide OP to the user U. For example, the related information Z is various information such as a character string expressing the pronunciation content of the guidance voice V, a character string translated from the language of the pronunciation content into another language, or a voice. In the first embodiment, distribution information Q for presenting various guidance sentences A (A1, A2,...) Prepared in advance related to the guidance voice V as related information Z to the user U is generated.

<Voice guidance system 100>
FIG. 2 is a configuration diagram of the voice guidance system 100 and the management device 10. As illustrated in FIG. 2, the voice guidance system 100 includes a distribution terminal 20, a sound collection device 22, an adder 24, and a sound emission device 26.

  The sound collection device 22 is an acoustic device (microphone) that collects ambient sounds. The guide OP selectively pronounces a plurality of sentences recorded in advance in an announcement book or the like to be referred to when the guidance voice V is sounded, for example, as the guidance voice V according to information to be guided. That is, the guidance voice V is basically a known content prepared in advance, not the content that the guide OP can arbitrarily determine.

  The sound collection device 22 collects the guidance voice V generated by the guider OP and generates a voice signal SG representing the time waveform of the guidance voice V. Note that an A / D converter that converts the audio signal SG generated by the sound collection device 22 from analog to digital is not shown for convenience.

  The sound signal SG generated by the sound collecting device 22 is supplied to the sound emitting device 26 via the adder 24 as an acoustic signal S1. The sound emitting device 26 is an acoustic device (speaker) that emits sound according to the acoustic signal S 1 supplied from the adder 24. For example, the guidance voice V represented by the voice signal SG is emitted from the sound emitting device 26 to the user U. The D / A converter that converts the acoustic signal S1 from digital to analog is not shown for convenience.

  As understood from the above description, the voice guidance system 100 connects the distribution terminal 20 and the adder 24 to an existing in-house broadcasting system that broadcasts the guidance voice V picked up by the sound pickup device 22 from the sound output device 26. Sound system. However, the form of the voice guidance system 100 is arbitrary. For example, each element of the distribution terminal 20, the sound collecting device 22, the adder 24, and the sound emitting device 26 are mounted on a single device, or the sound collecting device 22 is distributed. It can also be mounted on the terminal 20.

  The audio signal SG generated by the sound collection device 22 is branched from the path between the sound collection device 22 and the adder 24 and supplied to the distribution terminal 20. It is also possible to supply the audio signal SG to the distribution terminal 20 wirelessly.

  The distribution terminal 20 is a device for providing the terminal device 30 with distribution information Q corresponding to the guidance voice V of the audio signal SG supplied from the sound collection device 22, and is, for example, a portable type such as a mobile phone or a smartphone. Realized by a terminal device. As illustrated in FIG. 2, the distribution terminal 20 includes a control device 210 and a communication device 220.

  As illustrated in FIG. 2, the communication device 220 is a device that communicates with the management device 10 via the communication network 200. The communication device 220 transmits the audio signal SG to the management device 10 and receives the distribution information Q distributed from the management device 10 to the audio signal SG.

  The control device 210 is configured by a processing device such as a CPU (Central Processing Unit), for example, and controls the overall operation of the distribution terminal 20. Specifically, as illustrated in FIG. 2, the control device 210 executes a program stored in a known recording medium (not shown) such as a semiconductor recording medium or a magnetic recording medium, so that the guidance voice V can be handled. A plurality of functions (voice acquisition unit 212, signal processing unit 214) for acquiring and distributing the distributed information Q are realized.

  The voice acquisition unit 212 acquires the voice signal SG of the guidance voice V from the sound collection device 22 and transmits the voice signal SG from the communication device 220 to the management device 10 via the communication network 200. The management apparatus 10 receives the voice signal SG transmitted from the voice guidance system 100 and generates distribution information Q for presenting the user U with related information Z corresponding to the guidance voice V of the voice signal SG. The distribution information Q generated by the management device 10 is transmitted from the management device 10 to the voice guidance system 100. The communication device 220 receives the distribution information Q transmitted from the management device 10 from the communication network 200.

  The signal processing unit 214 generates an acoustic signal SQ including the distribution information Q received by the communication device 220 from the management device 10. The acoustic signal SQ contains the distribution information Q as an acoustic component in a predetermined frequency band. Specifically, the frequency band of the acoustic signal SQ is a frequency band in which sound emission by the sound emission device 26 and sound collection by the terminal device 30 can be performed, and the sound that the user U listens in a normal environment ( For example, it is included in the range of a frequency band (for example, 18 kHz or more and 20 kHz or less) exceeding the frequency band of sound such as guidance voice V) or musical sound. A known technique can be arbitrarily employed to generate the acoustic signal SQ by the signal processing unit 214. For example, the acoustic signal SQ is generated by frequency-modulating a carrier wave such as a sine wave having a predetermined frequency with the distribution information Q. A configuration or a configuration in which the acoustic signal SQ is generated by sequentially executing the spread modulation of the distribution information Q using a spread code and the frequency conversion using a carrier wave of a predetermined frequency may be employed.

  The adder 24 of the voice guidance system 100 generates the acoustic signal S1 by adding the acoustic signal SQ generated by the signal processing unit 214 and the voice signal SG generated by the sound collecting device 22. Therefore, the sound component of the distribution information Q is emitted from the sound emitting device 26 together with the guidance voice V that is generated by the guide OP. As understood from the above description, the sound emitting device 26 functions as an element that transmits the distribution information Q to the terminal device 30 by acoustic communication using sound (sound wave) as air vibration as a transmission medium. That is, the sound emitting device 26 that emits the guidance voice V collected by the sound collecting device 22 is used for transmission of the distribution information Q. Therefore, the configuration of the information generation system 1 is simplified compared to the configuration in which the sound emitting device 26 that is different from the sound emitting device 26 used for sounding the guidance voice V is used for sound emission of the sound indicating the distribution information Q. It is possible to

<Management device 10>
The management device 10 is a device that manages the distribution information Q transmitted to the voice guidance system 100, and includes a control device 110, a storage device 120, and a communication device 130 as illustrated in FIG. The management device 10 can be realized as a single device or a plurality of devices configured separately from each other. For example, the storage device 120 (cloud storage) may be installed separately from the management device 10 and the control device 110 may execute read / write with respect to the storage device 120 via the communication network 200, for example. That is, the storage device 120 can be omitted from the management device 10.

  The communication device 130 communicates with the distribution terminal 20 via the communication network 200. Specifically, the communication device 130 receives the audio signal SG transmitted from the distribution terminal 20 and transmits the distribution information Q generated by the control device 110 to the distribution terminal 20 in response to the audio signal SG.

  The control device 110 is configured by a processing device such as a CPU, for example, and executes a program stored in the storage device 120 to execute a plurality of elements (voices) that control generation of the distribution information Q as illustrated in FIG. It functions as an analysis unit 112, a character string specifying unit 114, and an information generation unit 116). A configuration in which a part of the function of the control device 110 is realized by a dedicated electronic circuit, or a configuration in which the function of the control device 110 is distributed to a plurality of devices may be employed.

  The voice analysis unit 112 in FIG. 2 generates a recognized character string L obtained by dividing a character string representing the pronunciation content of the guidance voice V into a plurality of parts. Specifically, the voice analysis unit 112 analyzes a result K (hereinafter referred to as “analysis result”) K obtained by analyzing the pronunciation content of the guidance voice V by voice recognition with respect to the voice signal SG received from the distribution terminal 20 by the communication device 130 in sentence units. Is divided into a plurality of recognized character strings L. FIG. 3 is an explanatory diagram of an operation of dividing the analysis result K of the guidance voice V into a plurality of recognized character strings L. As illustrated in FIG. 3, the analysis result K (K1, K2, K3) can be divided into a plurality of recognized character strings L in units of sentences. For voice recognition of the voice signal SG, a known technique such as a recognition technique using an acoustic model such as an HMM (Hidden Markov Model) and a language model indicating linguistic restrictions can be arbitrarily employed. For dividing the analysis result K, a known technique such as morphological analysis can be arbitrarily employed.

  The storage device 120 stores a program executed by the control device 110 and various data used by the control device 110. For example, a known recording medium such as a semiconductor recording medium or a magnetic recording medium or a combination of a plurality of types of recording media is arbitrarily employed as the storage device 120. As illustrated in FIG. 2, the storage device 120 stores a character string table TA and a guide table TB.

  FIG. 4 is a schematic diagram of the character string table TA. As illustrated in FIG. 4, the character string table TA is a data table in which a plurality of registered character strings X constituting different sentences assumed as the guidance voice V are registered. Each registered character string X is associated with identification information DX (DX1, DX2,...) For uniquely identifying the registered character string X. A plurality of registered character strings X (character strings expected to be pronounced by the guide OP) constituting each of a plurality of types of sentences recorded in the announcement book or the like are registered in the character string table TA.

  As described above, each of a plurality of sentences recorded in advance in an announcement book or the like is configured by combining a plurality of registered character strings X. The registered character string X may be common among a plurality of sentences. For example, the sentence “Thank you very much for coming today” is common between the sentence notifying the closing of the store and the sentence notifying the lost child as illustrated in FIG. As illustrated in FIG. 4, the registered character string X of the character string table TA is set to one sentence unit, so that it is not necessary to register the common registered character string X in the character string table TA in duplicate. Therefore, the data amount of the character string table TA can be reduced as compared with the configuration in which sentences of a plurality of different guidance voices V are individually registered in the character string table TA.

  The character string specifying unit 114 in FIG. 2 is similar to the recognized character string L among the plurality of registered character strings X registered in the character string table TA for each of the plurality of recognized character strings L analyzed by the voice analysis unit 112. The registered character string X to be specified is specified. Specifically, the character string specifying unit 114 calculates an index of similarity to the recognized character string L (hereinafter referred to as “similar index”) for each of the plurality of registered character strings X in the character string table TA, and Among the registered character strings X, one registered character string X (ie, the registered character string X most similar to the recognized character string L) having the maximum similarity indicated by the similarity index is specified.

  As described above, since the guide OP pronounces sentences (sentences composed of a combination of a plurality of registered character strings X) recorded in advance in an announcement book or the like to be referred to at the time of pronunciation, ideally a voice analysis unit Each of the plurality of recognized character strings L that 112 generates from the speech signal SG coincides with any registered character string X registered in the character string table TA. However, actually, erroneous recognition may occur in the analysis by the voice analysis unit 112 due to the utterance characteristics (feces) unique to each guider OP, background noise in the commercial facility 300, and the like. Therefore, each recognized character string L and registered character string X are similar to each other, but may not necessarily match. For example, the recognized character string L actually generated by the voice analysis unit 112 is exemplified in FIG. 3 even if the guider OP pronounces a sentence including the registered character string X4 “Please come home carefully”. As can be seen, it can be an analysis result K1 including a recognized character string L that is similar to the registered character string X but does not completely match, such as “... In the first embodiment, since the registered character string X similar to the recognized character string L is specified, as illustrated in FIG. 4, the registered character string X4 that does not include the influence of the misrecognition of voice recognition can be specified.

  The type of the similarity index is arbitrary, but a known index such as an edit distance (Levenstein distance) for evaluating the similarity between character strings can be arbitrarily adopted as the similarity index. The specification of the registered character string X by the character string specifying unit 114 is also referred to as a process of correcting each of the plurality of recognized character strings L generated by the speech analysis unit 112 to a registered character string X similar to the recognized character string L. obtain. The character string specifying unit 114 acquires the identification information DX from the character string table TA for each of the plurality of registered character strings X specified by the above procedure.

  FIG. 5 is a schematic diagram of the guide table TB. As illustrated in FIG. 5, the guide table TB includes a plurality of pieces of correspondence information C corresponding to different sentences (combination of a plurality of sentences) expected to be pronounced by the guide OP. Correspondence information C corresponding to an arbitrary sentence specifies a combination of identification information DX of a plurality of registered character strings X constituting the sentence. For example, correspondence information C2 corresponding to the sentence "Thank you for coming today. There is an urgent notice. Your car's headlight is still on." A combination of identification information DX1 of registered character string X1, identification information DX3 of registered character string X3 of the second sentence, and identification information DX6 of registered character string X6 of the third sentence is designated. Identification information DZ (DZ1, DZ2,...) Is associated with each of the plurality of correspondence information C in the guide table TB. The identification information DZ designates various guidance sentences A prepared in advance related to the guidance voice V. The guidance sentence A is an appropriate sentence recorded in advance in an announcement book or the like. Specifically, the identification information DZ corresponding to any one piece of correspondence information C designates a guide sentence A composed of a plurality of registered character strings X in which the correspondence information C designates the identification information DX. Note that it is also possible to have identification information DZ for each predetermined unit (for example, sentence) constituting the guidance sentence A, and a plurality of identification information DZ can designate one guidance sentence A.

  The information generation unit 116 in FIG. 2 corresponds to a combination of a plurality of registered character strings X acquired by the character string specifying unit 114 among a plurality of pieces of correspondence information C that specify different combinations of the plurality of registered character strings X. The distribution information Q for the terminal device 30 to present to the user U the related information Z corresponding to the information C is generated. Specifically, the information generation unit 116 specifies correspondence information C that specifies a combination that is similar to or matches the combination of the plurality of pieces of identification information DX acquired by the character string specification unit 114 among the plurality of pieces of correspondence information C. The identification information DZ of the identified correspondence information C is acquired from the guidance table TB. In other words, the information generation unit 116 specifies correspondence information C that specifies the same combination as the plurality of identification information DX acquired by the character string specifying unit 114, and also sentences previously recorded in the announcement book or the like in sentence units. When the guide OP makes a mistake in pronunciation (for example, when the pronunciation of a necessary sentence is forgotten or an unnecessary sentence is pronounced), the information generation unit 116 resembles a plurality of pieces of identification information DX of the sentence. Correspondence information C specifying the combination to be identified is identified, and identification information DZ of the identified correspondence information C is acquired from the guidance table TB.

  For example, in the announcement book, the text “Thank you for coming today. There is an urgent notice. The car headlight remains on.” 4 is accurately pronounced, the character string specifying unit 114 acquires identification information DX1, DX3, and DX6 from the character string table TA, as illustrated in FIG. As illustrated in FIG. 5, since the identification information DX1, DX3, and DX6 acquired by the character string specifying unit 114 matches the identification information DX1, DX3, and DX6 specified by the correspondence information C2, the information generation unit 116 The identification information DZ2 of the correspondence information C2 is acquired from the guidance table TB. On the other hand, if there is an urgent notice, thank you for coming today. The car headlight is still on. However, the information generation unit 116 can acquire the identification information DZ2 of the correspondence information C2 from the guidance table TB. Also, the message “I have an urgent notice” is overwritten. “Thank you for coming today. Your car ’s headlight is still on.” In the case of pronunciation, as illustrated in FIG. 4, the character string specifying unit 114 acquires the identification information DX1 and DX6 from the character string table TA. When the correspondence information C including both the identification information DX1 and DX6 includes only the correspondence information C2 specifying the identification information DX1, DX3, and DX6, the information generating unit 116 identifies the identification information of the correspondence information C2 from the guidance table TB. Get DZ2. Therefore, it is possible to obtain the identification information DZ2 associated with the appropriate guide sentence A2.

  The information generation unit 116 in FIG. 2 generates distribution information Q including the identification information DZ acquired by the above procedure, and transmits the distribution information Q to the distribution terminal 20 via the communication device 130.

  As understood from the above description, the distribution information Q for presenting the user U with the related information Z indicating the guidance sentence A prepared independently of the combination of the identification information DX designated by the correspondence information C is generated. It is possible. As described above, the guidance sentence A is assumed to be a sentence recorded in advance in an announcement book or the like (that is, a sentence composed of a combination of registered character strings X specified by the correspondence information C). It is also possible to use a sentence different from the sentence recorded in the guide sentence A. For example, the text recorded in the announcement book etc. "Thank you for coming today. (= X1) It will be closed soon. (= X2) Please be careful. ) "A guide sentence A1 corresponding to the identification information DZ1 of the correspondence information C1 indicating the combination of the three registered character strings X that constitute") "content such as" Please be careful, as it will be closing time soon. " It is also possible to make it. That is, a sentence different from the sentence recorded in advance in the announcement book or the like can be used as the guide sentence A.

  In the voice guidance system 100 of FIG. 2, the acoustic component of the distribution information Q and the guidance voice V received by the communication device 220 from the management device 10 are emitted from the sound emitting device 26. Since the delivery information Q is generated by the process performed by the control device 110 after the utterance of the guidance voice V by the guider OP, the delivery information is delayed from the sound emission of the guidance voice V from the sound emitting device 26. The Q acoustic component is emitted.

<Terminal device 30>
FIG. 6 is a configuration diagram of the terminal device 30. As illustrated in FIG. 6, the terminal device 30 includes a sound collection device 310, a control device 320, a storage device 330, and a presentation device 340.

  The storage device 330 stores a program executed by the control device 320 and various data used by the control device 320. The storage device 330 stores a presentation table TC as illustrated in FIG. In the presentation table TC of the first embodiment, identification information DZ (DZ1, DZ2,...) Is associated with each of a plurality of guidance sentences A (A1, A2,...) Corresponding to different guidance voices V. It is a data table. For example, a known recording medium such as a semiconductor recording medium or a magnetic recording medium or a combination of a plurality of types of recording media is arbitrarily employed as the storage device 330.

  The sound collecting device 310 is an acoustic device (microphone) that picks up surrounding sounds, and collects sound emitted from the sound emitting device 26 of the distribution terminal 20 to generate an acoustic signal S2. The acoustic signal S2 contains the acoustic component (acoustic signal SQ) of the distribution information Q. The A / D converter that converts the acoustic signal S2 generated by the sound pickup device 310 from analog to digital is not shown for convenience.

  The control device 320 is constituted by a processing device such as a CPU, for example, and is designated by the distribution information Q among a plurality of guidance sentences A registered in the presentation table TC by executing a program stored in the storage device 330. A plurality of functions (information extraction unit 322 and information management unit 324) for presenting the guidance sentence A to the user U as related information Z are realized.

  The information extraction unit 322 extracts the distribution information Q by demodulating the acoustic signal S2 generated by the sound collection device 310. Specifically, the information extraction unit 322 emphasizes the band component of the frequency band including the distribution information Q in the acoustic signal S2 with, for example, a band pass filter, and uses the spreading code used for the spread modulation of the distribution information Q as the coefficient. The distribution information Q is extracted by passing the matched filter.

  The information management unit 324 selects the guide sentence A corresponding to the identification information DZ included in the distribution information Q extracted by the information extraction unit 322 from among the plurality of guide sentences A in the presentation table TC. That is, the guidance sentence A related to the guidance voice V pronounced by the guider OP is selected as the related information Z.

  The presentation device 340 presents the guidance sentence A selected by the information management unit 324 to the user U of the terminal device 30 as related information Z. The presentation device 340 is a display device (for example, a liquid crystal display panel) that displays the guidance sentence A indicated by the related information Z. As understood from the above description, the guidance sentence A corresponding to the utterance content of the guidance voice V pronounced by the guide OP is presented to the user U by the presentation device 340 as the related information Z. The user U visually recognizes the related information Z presented on the presentation device 340, so that the related information Z related to the guidance voice V generated by the guide OP and emitted from the sound emitting device 26 is visually displayed. It is possible to confirm.

  FIG. 7 is an explanatory diagram of the overall operation of the information generation system 1. When the guider OP generates a guidance voice V corresponding to a sentence recorded in advance in an announcement book or the like, the sound collection device 22 of the voice guidance system 100 collects the guidance voice V and generates a voice signal SG ( SA1). The sound signal SG generated by the sound collection device 22 is output from the sound collection device 22 to the sound emission device 26 and emitted (SA2), while being acquired by the sound acquisition unit 212 and transmitted from the communication device 220 to the communication network 200. It is transmitted (SA3).

  When the communication device 130 receives the voice signal SG transmitted from the distribution terminal 20 from the communication network 200, the voice analysis unit 112 of the management device 10 generates the pronunciation content of the guidance voice V by voice recognition for the voice signal SG of the guidance voice V. A plurality of recognized character strings L to be represented are generated (SA4). The character string specifying unit 114 specifies a registered character string X similar to each of the plurality of recognized character strings L generated by the speech analysis unit 112 from the plurality of registered character strings X in the character string table TA (SA5). That is, the recognized character string L is corrected to the registered character string X in which erroneous recognition by the voice analysis unit 112 is eliminated. The information generation unit 116 identifies the correspondence information C corresponding to the combination of the plurality of registered character strings X identified by the character string identification unit 114 from the plurality of correspondence information C in the guidance table TB (SA6), and the identified correspondence information Delivery information Q including C identification information DZ is generated (SA7). The communication device 130 transmits the distribution information Q generated by the information generation unit 116 to the distribution terminal 20 (SA8).

  When the communication device 220 receives the distribution information Q transmitted from the management device 10, the signal processing unit 214 of the distribution terminal 20 generates an acoustic signal SQ containing the distribution information Q as an acoustic component (SA9). The adder 24 of the distribution terminal 20 generates the acoustic signal S1 by adding the audio signal SG generated by the sound collection device 22 and the acoustic signal SQ generated by the signal processing unit 214 (SA10). The sound emitting device 26 emits sound according to the acoustic signal S1 (SA11). That is, the guidance voice V picked up by the sound pickup device 22 and the acoustic component of the distribution information Q are emitted from the sound emission device 26.

  The sound collecting device 310 of the terminal device 30 collects the sound emitted from the sound emitting device 26 and generates an acoustic signal S2 (SA12). The information extraction unit 322 extracts the distribution information Q by demodulating the acoustic signal S2 generated by the sound collection device 310 (SA13). The information management unit 324 selects the guidance sentence A corresponding to the identification information DZ included in the distribution information Q extracted by the information extraction unit 322 from the presentation table TC. The presentation device 340 visually presents to the user U by displaying the character string indicated by the guidance sentence A selected by the information management unit 324 as the related information Z (SA14).

  As understood from the above description, in the first embodiment, among the plurality of registered character strings X, the registered character string X similar to the recognized character string L analyzed by the speech recognition for the guidance voice V is specified. It is possible to generate distribution information Q for presenting related information Z that does not include the influence of misrecognition of voice recognition to the user U. Further, since the distribution information Q for presenting the user U with the related information Z corresponding to the correspondence information C designating different combinations of the plurality of registered character strings X is generated, the guide guides for example in sentence units. A combination of a plurality of registered character strings X identified from the analysis result K of the guidance voice V even when the pronunciation of the voice V is wrong (for example, when a necessary sentence is forgotten or an unnecessary sentence is pronounced) It is possible to generate distribution information Q for presenting relevant information Z corresponding to the correspondence information C corresponding to the user U to the user U.

Second Embodiment
A second embodiment of the present invention will be described. Regarding the elements whose functions and functions are the same as those of the first embodiment in each aspect exemplified below, the detailed description of each is appropriately omitted by using the reference numerals used in the description of the first embodiment.

  The information generation unit 116 according to the first embodiment uses the guidance sentence A corresponding to the identification information DZ of the correspondence information C that specifies a combination of a plurality of registered character strings X specified by the character string specifying unit 114 as related information Z. Distribution information Q to be presented to U is generated. The information generation unit 116 according to the second embodiment displays the related information Z indicating the plurality of registered character strings X specified by the correspondence information C corresponding to the combination of the plurality of registered character strings X specified by the character string specifying unit 114. 30 generates distribution information Q for presentation to the user U.

  FIG. 8 is a schematic diagram of a guide table TB according to the second embodiment. As illustrated in FIG. 8, the guide table TB according to the second embodiment includes a plurality of pieces of correspondence information C corresponding to different sentences assumed to be pronounced by the guide OP. Correspondence information C corresponding to an arbitrary sentence specifies a combination of identification information DX of a plurality of registered character strings X constituting the sentence. For example, correspondence information C2 corresponding to the sentence "Thank you for coming today. There is an urgent notice. Your car's headlight is still on." A combination of identification information DX1 of registered character string X1, identification information DX3 of registered character string X3 of the second sentence, and identification information DX6 of registered character string X6 of the third sentence is designated. Compared with the guide table TB according to the first embodiment of FIG. 5, the guide table TB according to the second embodiment does not have the identification information DZ corresponding to any one correspondence information C.

  The information generation unit 116 specifies correspondence information C corresponding to the plurality of pieces of identification information DX acquired by the character string specification unit 114 from the plurality of pieces of correspondence information C, as in the first embodiment. The information generation unit 116 acquires a plurality of pieces of identification information DX designated by the identified correspondence information C from the guidance table TB, and generates distribution information Q including the plurality of pieces of identification information DX. For example, when the correspondence information C2 is specified, the information generation unit 116 generates the distribution information Q including the identification information DX1, the identification information DX3, and the identification information DX6, and transmits the distribution information Q to the distribution terminal 20 via the communication device 130. Send.

  In the terminal device 30, the sound collecting device 310 picks up the sound emitted from the sound emitting device 26 of the distribution terminal 20 to generate the sound signal S 2, and the information extraction unit 322 extracts the distribution information Q from the sound signal S 2. The configuration and operation are the same as those in the first embodiment.

  In the second embodiment, the presentation table TC stored in the storage device 330 of the terminal device 30 is data in which the identification information DX is associated with each of the plurality of registered character strings X, similarly to the character string table TA illustrated in FIG. It is a table. The information management unit 324 acquires the registered character string X of each identification information DX specified by the distribution information Q extracted by the information extraction unit 322 from the presentation table TC, and the presentation device 340 includes a plurality of information selected by the information management unit 324. The registered character string X is presented to the user U as related information Z.

  In the above example, the presentation table TC has the same contents as the character string table TA, but the presentation table TC may have different contents from the character string table TA. For example, the registered character string X5 corresponding to the identification information DX5 of the character string table TA is “I will give you a notification of lost child”, but the character string corresponding to the identification information DX5 of the presentation table TC is “Notice of lost child”. It is also possible. The information management unit 324 selects a registered character string X indicated by each of the plurality of identification information DX included in the distribution information Q from among the plurality of identification information DX in the presentation table TC, and selects the selected plurality of registered character strings X. Is generated as related information Z. The presentation device 340 presents the related information Z generated by the information management unit 324 to the user U of the terminal device 30.

  In the second embodiment, the same effect as in the first embodiment is realized. In the second embodiment, distribution information Q for presenting the user U with related information Z indicating a plurality of registered character strings X designated by the correspondence information C is generated. Compared with the guidance table TB according to the first embodiment of FIG. 5, as described above, the guidance table TB according to the second embodiment does not have the identification information DZ corresponding to any one correspondence information C. . Therefore, the distribution information Q for presenting the user U with the related information Z indicating the guidance sentence A of the correspondence information C corresponding to the combination of the plurality of registered character strings X specified by the character string specifying unit 114 is generated. Compared with the configuration of the embodiment, there is an advantage that information for uniquely designating the guidance sentence A (that is, identification information DZ) is not required. On the other hand, in the first embodiment, distribution for presenting the user U with related information Z indicating the guidance sentence A of the corresponding information C corresponding to the combination of the plurality of registered character strings X specified by the character string specifying unit 114. Information Q is generated. Accordingly, it is possible to generate the distribution information Q for presenting the user U with the related information Z indicating the guidance sentence A prepared independently of the combination of the registered character strings X designated by the correspondence information C.

<Third Embodiment>
The correspondence information C of the second embodiment designates a combination of a plurality of registered character strings X, and the information generation unit 116 provides related information Z indicating the plurality of registered character strings X designated by the correspondence information C to the terminal device 30. The distribution information Q to be presented to the user U is generated. The correspondence information C of the third embodiment designates a combination of a plurality of registered character strings X and the order of the plurality of registered character strings X related to the combination, and the information generation unit 116 designates the order designated by the correspondence information C. The distribution information Q for the terminal device 30 to present to the user U the related information Z indicating the plurality of registered character strings X arranged in the above is generated.

  The guide table TB according to the third embodiment includes a plurality of pieces of correspondence information C corresponding to different sentences that are supposed to be pronounced by the guide OP. Correspondence information C corresponding to an arbitrary sentence is composed of a combination of identification information DX of a plurality of registered character strings X constituting the sentence and a registered character string X indicated by each of the plurality of identification information DX. Specify the order to perform. Specifically, any one piece of correspondence information C is information in which identification information DX of each registered character string X is arranged in a regular order of a plurality of registered character strings X constituting a sentence corresponding to the correspondence information C. It is. However, the method of designating the order of the plurality of registered character strings X is not limited to the above examples. For example, it is possible to add information indicating the order of the registered character strings X in the combination to the identification information DX of each of the plurality of registered character strings X related to the combination designated by the correspondence information C.

  The information generation unit 116 according to the third embodiment generates the distribution information Q by arranging a plurality of pieces of identification information DX designated by the correspondence information C in the order designated by the correspondence information C. For example, “There is an urgent notice. (= DX3) Thank you very much for coming today. (= DX1) Your car headlight remains on. (= DX6)” If the correspondence information C specifies the order of “identification information DX1: No.1, identification information DX3: No.2, identification information DX6: No.3” Thank you. (= DX1) There is an urgent notice. (= DX3) The car headlight is still on. (= DX6) Distribution information Q for presentation to U is generated.

  In the third embodiment, the same effect as in the second embodiment is realized. Also, distribution information Q for the terminal device 30 to present to the user U related information Z indicating a plurality of registered character strings X arranged in the order specified by the correspondence information C is generated. Therefore, regardless of the order of pronunciation of the plurality of sentences constituting the guidance voice V, the terminal device 30 displays the related information Z indicating the plurality of registered character strings X in which the correspondence information C is arranged in a predetermined order. It is possible to generate distribution information Q for presentation to the user.

<Fourth embodiment>
In the first to third embodiments, the distribution information Q instructing the presentation of the related information Z is transmitted from the information generation system 1 to the terminal device 30. In the fourth embodiment, the information generation system 1 generates related information Z corresponding to the guidance voice V and provides it to the user U. The operations of the voice analysis unit 112 and the character string specifying unit 114 and the operation of the information generating unit 116 specifying the correspondence information C are the same as in the above-described embodiments. Therefore, similarly to each of the above-described embodiments, it is possible to present the user U with appropriate related information Z corresponding to the correspondence information C, in which the influence of erroneous recognition of voice recognition is reduced.

  The information generation unit 116 according to the fourth embodiment generates, as related information Z, a character string obtained by translating the guidance sentence A indicated by the identification information DZ of the identified correspondence information C into another language. For example, a known machine translation can be arbitrarily adopted for the translation of the character string. The related information Z generated by the information generation unit 116 is transmitted to the distribution terminal 20 of the voice guidance system 100.

  The signal processing unit 214 of the distribution terminal 20 generates the acoustic signal SQ by voice synthesis using the related information Z. The acoustic signal SQ according to the fourth embodiment is a signal representing a voice that is produced by sounding a character string designated by the related information Z. For the generation of the acoustic signal SQ, known speech synthesis can be arbitrarily adopted. The acoustic signal SQ generated by the signal processing unit 214 is supplied to the sound emitting device 26 via the adder 24. Therefore, the sound that pronounces the character string specified by the related information Z is emitted from the sound emitting device 26. That is, following the sound emission of the guidance voice V pronounced by the guide OP, a voice obtained by translating the guidance voice V into another language is emitted from the sound emission device 26 to the user U. For example, when the translated sentence of the guide sentence A is generated as the related information Z, the voice of the translated sentence of the guide sentence A corresponding to the guide voice V is emitted following the guide voice V.

  In the above description, the configuration based on the first embodiment has been exemplified. However, it is also possible to apply the character string specification in the second embodiment or the third embodiment to the fourth embodiment. For example, the information generation unit 116 according to the fourth embodiment converts a character string obtained by translating a sentence in which a registered character string X indicated by each of a plurality of pieces of identification information DX specified by the identified correspondence information C into another language is related information Z. Generate as Accordingly, the voice of the translated sentence of the sentence in which the plurality of registered character strings X are combined is emitted from the sound emitting device 26 following the guidance voice V. Further, assuming the third embodiment, a character obtained by translating a sentence in which a registered character string X indicated by each of a plurality of identification information DX designated by the identified correspondence information C is arranged in the order designated by the correspondence information C into another language It is also possible for the information generating unit 116 to generate related information Z representing a column. In the above-described configuration, the translated speech of a sentence in which a plurality of registered character strings X are arranged in the order designated by the correspondence information C is emitted from the sound emitting device 26 following the guidance voice V.

As understood from the above description, the information generation system 1 of the fourth embodiment is a system that generates (and provides to the user U) related information Z related to the guidance voice V, and recognizes the voice for the guidance voice V. A character string specifying unit 114 that specifies the registered character string X similar to the recognized character string L among the plurality of registered character strings X representing different pronunciation contents for each of the plurality of recognized character strings L analyzed in Related information Z corresponding to the correspondence information C corresponding to the combination of the plurality of registered character strings X identified by the character string identifying unit 114 among the plurality of correspondence information C designating different combinations of the plurality of registered character strings X And an information generation unit 116 to be generated. A typical example of the related information Z corresponding to the correspondence information C is a sentence composed of the translated sentence of the guidance sentence A exemplified in the first embodiment and a plurality of registered character strings X indicated by the correspondence information C (second embodiment). Or a translated sentence of the third embodiment). In addition, although the structure which emits the sound of the character string which the related information Z shows from the sound emission apparatus 26 was illustrated in the above description, the output method of the related information Z is not limited to the above illustration. For example, a character string indicated by the related information Z can be displayed on the display device.
<Modification>
Each aspect illustrated above can be variously modified. Specific modifications are exemplified below. Two or more modes arbitrarily selected from the following examples can be appropriately combined within a range that does not contradict each other.

(1) In the first to third embodiments, the information generating unit 116 uses character string identification information DZ (identification information DX in the second and third embodiments) indicating the pronunciation of the guidance voice V. Although generated as distribution information Q, the contents of distribution information Q are not limited to the above examples. For example, the presentation device 340 can display a character string indicating the pronunciation content of the guidance voice V or a character string obtained by translating the pronunciation content of the guidance voice V into another language. However, the method of presentation to the user U is not limited to display. For example, a sound emitting device that emits the guidance sentence A or the registered character string X designated by the distribution information Q as sound can be used as the presentation device 340. As understood from the above examples, the distribution information Q is comprehensively expressed as information on the guidance voice V.

(2) In each of the above-described forms, the information generation unit 116 specifies the correspondence information C and then generates the distribution information Q (related information Z in the fourth embodiment). It is also possible to consider the case where it cannot be specified. For example, on the basis of the configuration of the first to third embodiments, when the information generating unit 116 cannot identify the correspondence information C, the configuration or distribution information Q is generated with the analysis result K of the guidance voice V as the distribution information Q. A configuration without this is also possible. Further, on the basis of the configuration of the fourth embodiment, when the information generation unit 116 cannot identify the correspondence information C, a configuration in which the translated information of the character string of the analysis result K of the guidance voice V is the related information Z and the related information Z are A configuration without generation is also possible.

(3) In each of the above-described forms, a sentence constituting a sentence recorded in an announcement book or the like is a registered character string X. However, a sentence composed of a fixed phrase provided with an insertion section and an insertion phrase is a registered character string. X can also be used. For example, a registered character string X is a sentence in which an insertion phrase similar to a portion corresponding to the insertion section of the analysis result K is inserted into a fixed phrase similar to the analysis result K. In the above configuration, one fixed phrase is shared over a plurality of analysis results K (a character string common to a plurality of guidance voices V need not be included in each registered character string X of the character string table TA). Therefore, there is an advantage that the data amount of the character string table TA is reduced.

(4) In each of the above-described embodiments, the configuration in which the management device 10 includes the voice analysis unit 112, the character string specifying unit 114, and the information generation unit 116 is illustrated. However, some or all of the functions of the management device 10 are voiced. It can also be installed in the guidance system 100. For example, in the configuration in which the voice analysis unit 112, the character string specifying unit 114, and the information generation unit 116 are mounted on the distribution terminal 20 on the basis of the configuration of the first to third embodiments, the analysis of the voice signal SG (voice The analysis unit 112), identification of the registered character string X (character string identification unit 114), and generation of distribution information Q (information generation unit 116) are executed by the distribution terminal 20, and the distribution information Q is output from the sound emitting device 26. To the terminal device 30. In this configuration, since communication between the voice guidance system 100 and the management device 10 is unnecessary, it is possible to provide the distribution information Q to the terminal device 30 even in an environment where communication using the communication network 200 is not possible. On the other hand, in the configuration in which the speech analysis unit 112, the character string specifying unit 114, and the information generation unit 116 are mounted on the distribution terminal 20 on the basis of the configuration of the fourth embodiment, the analysis of the speech signal SG (speech analysis unit 112) and The identification of the registered character string X (character string specifying unit 114) and the generation of the related information Z (information generating unit 116) are executed by the distribution terminal 20, and the related information Z is emitted from the sound emitting device 26. .

(5) In the first embodiment to the third embodiment, the storage device 330 of the terminal device 30 stores the presentation table TC, but the storage destination of the presentation table TC is not limited to the above examples. For example, the presentation table TC can be stored in a distribution server device that communicates with the terminal device 30 via a communication network such as a mobile communication network or the Internet. The terminal device 30 transmits an information request specifying the identification information DZ (identification information DX in the second and third embodiments) included in the distribution information Q to the distribution server device. The guidance sentence A corresponding to the specified identification information DZ is transmitted to the requesting terminal device 30. The presentation device 340 of the terminal device 30 presents the guidance sentence A received from the distribution server device to the user U. As understood from the above description, it is not essential to store the presentation table TC in the storage device 330 of the terminal device 30.

(6) In each of the above embodiments, the speech analysis unit 112 divides the analysis result K of the guidance voice V by speech recognition into sentence units, but the unit for dividing the analysis result K is not limited to the above examples. For example, the analysis result K can be divided into phrases or words. The analysis result K may be divided in units of a predetermined number of sentences.

(7) In each above-mentioned form, although information generation system 1 was used in commercial facilities 300, such as a shopping mall, the place which uses information generation system 1 is not limited to the above illustration. For example, the information generation system 1 can be used for guidance of transportation facilities such as buses and trains.

(8) In the first to third embodiments, acoustic communication is used for distributing the distribution information Q to the terminal device 30. However, the communication method for distributing the distribution information Q to the terminal device 30 is limited to the above examples. Not. For example, the distribution information Q can be distributed to the terminal device 30 by wireless communication (for example, short-range wireless communication) using infrared rays or radio waves.

(9) In the second embodiment, the correspondence information C designates a combination of a plurality of registered character strings X, but the information designated by the correspondence information C is not limited to the above examples. For example, among the plurality of pieces of identification information DX indicated by the correspondence information C, the identification information DX of the registered character string X that is not included in the distribution information Q (that is, not presented to the user U) can be designated. For example, “Thank you for coming today. (= DX1) There is an urgent notice. (= DX3) The car headlight is still on. (= DX6)” When the operator OP pronounces, if the correspondence information C2 indicating the identification information DX1, DX3 and DX6 specifies “identification information DX1: non-presentation”, “There is an urgent notice. (= DX3) The distribution information Q for the terminal device 30 to present to the user U the related information Z (= DX6) ”can be generated. According to the above configuration, a sentence that is desired to be emitted as the guidance voice V, but a sentence that is not desired to be distributed to the terminal device 30 (for example, a sentence including personal information or a less important sentence) is distributed to the distribution information Q. Can be excluded.

(10) The information generation system 1 exemplified in each of the above embodiments is realized by the cooperation of the control device 110 of the management device 10 and a program as illustrated in the above embodiments. For example, the program corresponding to the first to third embodiments has the distribution information Q transmitted to the terminal device 30 in order for the terminal device 30 to present the related information Z related to the guidance voice V to the user U. For each of a plurality of recognized character strings L analyzed by voice recognition with respect to the guidance voice V, and similar to the recognized character string L among a plurality of registered character strings X representing different pronunciation contents Among the plurality of pieces of correspondence information C that specify different combinations of a plurality of registered character strings X, a plurality of registered character strings X specified by the character string specifying unit 114 are specified. The computer is caused to function as the information generation unit 116 that generates the distribution information Q for the terminal device 30 to present to the user U the related information Z corresponding to the correspondence information C corresponding to the combination. The program corresponding to the fourth embodiment is a program for generating (and providing to the user U) related information Z related to the guidance voice V, and a plurality of recognitions analyzed by voice recognition for the guidance voice V. For each of the character strings L, a character string specifying unit 114 for specifying the registered character string X similar to the recognized character string L among the plurality of registered character strings X representing different pronunciation contents, and the plurality of registered character strings X Information generating unit 116 that generates related information Z corresponding to the corresponding information C corresponding to the combination of the plurality of registered character strings X specified by the character string specifying unit 114 among the plurality of corresponding information C designating different combinations of As a computer to function. The programs exemplified above can be provided in a form stored in a computer-readable recording medium and installed in the computer. The recording medium is, for example, a non-transitory recording medium, and an optical recording medium (optical disk) such as a CD-ROM is a good example, but a known arbitrary one such as a semiconductor recording medium or a magnetic recording medium This type of recording medium can be included. It is also possible to distribute the program to a computer in the form of distribution via a communication network.

(11) The present invention is also specified as an operation method (information generation method) of the information generation system 1 according to each of the above-described embodiments. For example, in the information generation method corresponding to the first to third embodiments, the distribution transmitted to the terminal device 30 so that the terminal device 30 presents the related information Z related to the guidance voice V to the user U. A method of generating information Q, which is similar to a recognized character string L among a plurality of registered character strings X representing different pronunciation contents for each of a plurality of recognized character strings L analyzed by voice recognition for the guidance voice V In accordance with the correspondence information C corresponding to the specified combination of the plurality of registered character strings X among the plurality of correspondence information C specifying the registered character string X to be specified and specifying different combinations of the plurality of registered character strings X Distribution information Q for the terminal device 30 to present the related information Z to the user U is generated. The information generation method corresponding to the fourth embodiment is a method for generating (and providing to the user U) related information Z related to the guidance voice V, and a plurality of information analyzed by voice recognition with respect to the guidance voice V. For each of the recognized character strings L, the registered character string X similar to the recognized character string L is specified from among the plurality of registered character strings X representing different pronunciation contents, and different combinations of the plurality of registered character strings X are determined. Of the plurality of correspondence information C to be specified, the related information Z corresponding to the correspondence information C corresponding to the specified combination of the plurality of registered character strings X is generated.

DESCRIPTION OF SYMBOLS 1 ... Information generation system, 10 ... Management apparatus, 20 ... Distribution terminal, 22 ... Sound collection apparatus, 24 ... Adder, 26 ... Sound emission apparatus, 30 ... Terminal apparatus, 100 ... Voice guidance System 110 110 Control device 112 Speech analysis unit 114 Character string specifying unit 116 Information generation unit 120 Storage unit 130 Communication device 200 Communication network 210 ... Control device, 212 ... Audio acquisition unit, 214 ... Signal processing unit, 220 ... Communication device, 300 ... Commercial facility, 310 ... Sound pickup device, 320 ... Control device, 322 ... Information extraction unit, 324 ... Information management unit, 330 ... Storage device, 340 ... Presentation device

Claims (5)

A system for generating distribution information transmitted to a terminal device in order for the terminal device to present related information related to the guidance voice to the user,
For each of a plurality of recognized character strings analyzed by voice recognition with respect to the guidance voice, a character string specifying unit for specifying the registered character string similar to the recognized character string among a plurality of registered character strings representing different pronunciation contents When,
Of the plurality of pieces of correspondence information designating different combinations of the plurality of registered character strings, the terminal device displays the related information according to the correspondence information corresponding to the combination of the plurality of registered character strings specified by the character string specifying unit. An information generation system comprising: an information generation unit that generates the distribution information to be presented to a user. The information generation unit includes related information indicating a guidance sentence of correspondence information corresponding to a combination of a plurality of registered character strings identified by the character string identification unit among a plurality of guidance sentences prepared for each correspondence information. The information generation system according to claim 1, wherein the terminal device generates the distribution information to be presented to the user. The information generation unit is configured so that the terminal device presents to the user related information indicating a plurality of registered character strings specified by correspondence information corresponding to a combination of a plurality of registered character strings specified by the character string specifying unit. The information generation system according to claim 1, wherein the distribution information is generated. The correspondence information specifies a combination of a plurality of registered character strings and an order of the plurality of registered character strings related to the combination,
The information generation unit generates the distribution information for the terminal device to present related information indicating a plurality of registered character strings arranged in an order designated by the correspondence information to the user. Generation system. 5. The information generation system according to claim 1, further comprising a sound emitting unit that emits the guidance voice and sound indicating the distribution information.

Images