JP5315873B2 - Search device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a route search device recognizing a station name input by voice with high accuracy. <P>SOLUTION: The route search device applies weight of enhancing the priority of an area name, every time an area related to a user is derived from address data of an address book ADR and mail data of a mail box MB, and searches for the station name corresponding to a voice-recognized word from a route search database 15, when the user inputs the station name (getting-on and getting-off station names) by voice. The route search device applies weight of enhancing the priority of the corresponding area name to the station name existing in the area name related to the user (step SE6) when a plurality of station names are obtained, and displays the plurality of getting-on station names on a route search screen in the order of the priority of the weighted area name (step SE8). The station name existing in the area name related to the user is placed at a higher rank in selection candidate, and false recognition and troublesomeness are eliminated thereby to recognize the voice-input station name with high accuracy. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

The present invention relates to a retrieval apparatus and program searching voice entered name.

In response to the input of the boarding station name and the boarding station name, a route search device that searches for the optimal railway line from the boarding station to the boarding station and presents it to the user has been put to practical use. About this type of device (system) For example, it is disclosed in Patent Document 1. Normally, in the route search device, the boarding station name and the getting-off station name are input by a user's keyboard operation, but a voice input method has also been devised in order to simplify the input operation.

  The problem with the voice input method is erroneous recognition of the input voice. In particular, there are many very similar names (for example, “Kami-Fukuoka” and “Fukuoka”) as station names, and if the input speech is not accurately recognized, all the station names with similar names become input candidates. In that case, the user must search for a desired station name from among them, and as a result, further labor is required.

  By the way, in the field of speech recognition, various methods for avoiding misrecognition of speech input have been developed. For example, in the speech recognition apparatus disclosed in Patent Document 2, a dictionary database having a plurality of dictionary patterns corresponding to each region is provided, and the dictionary pattern of the region specified based on the current position obtained by GPS positioning is stored from the dictionary database. By searching and recognizing speech using a corresponding dictionary pattern, highly accurate speech recognition is realized regardless of the region used.

  If such a speech recognition technology is applied to a route search device of a speech input method, it is possible to eliminate the adverse effect that the input speech is not accurately recognized and all station names with similar names become input candidates. In other words, it is rare that the route to be searched is in a different area far from the area including the current position, so if the station name entered by voice is recognized as a station name in the area far from the current position, By excluding station names from input candidates, it is possible to avoid that all station names with similar names become input candidates.

Japanese Patent Laid-Open No. 2005-271691 JP-A-8-179790

However, as described above, when the station name in the area far from the current position is excluded from the input candidates, when searching for the route in the area far from the current position, the station name that should be the input candidate is excluded. This can lead to misrecognition and a search effort. In other words, there is a problem that a station name inputted by voice cannot be recognized with high accuracy.

An object of the present invention is to make it possible to search a name inputted by voice in accordance with a user's intention .

The invention according to claim 1 searches for user data storage means for storing a plurality of user data containing contents capable of deriving a name related to the user, and a name that matches the result of recognizing the input voice. If a search unit, the searching means searches a plurality of candidate name, based on the respective contents of a plurality of user data stored in the user data storage unit, obtains the weighting priority of each candidate names are those search Weight display means, and name display means for rearranging and displaying a plurality of candidate names searched by the search means based on the priority of the candidate names weighted by the weighting means, and the user The data storage means includes a first storage means for storing a plurality of first type user data including address data registered in response to a user operation, and a plurality of second types. Includes a second storage means for storing user data, wherein the weighting means, prior to retrieval by the retrieval means, the address data included in the plurality of first type user data stored in said first storage means based on obtains the priority of each of a plurality of names, further changes based on the presence or absence of the name of the priority for each name, included in the plurality of second type user data stored in the second storage means doing, the previously stored Sakido Yu of each name, a search for a plurality of candidate name in the search means, the weighting of the retrieved each candidate name was, corresponding to the respective candidate name And based on the priority of each name stored in advance .

In the present invention, a name inputted by voice can be searched according to the user's intention .

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
A. Constitution
(1) Appearance
FIG. 1 is an external view showing an external appearance of a route search apparatus 100 according to the first embodiment. The route search apparatus 100 shown in this figure is configured with a casing of a size that can be carried by a user, and an operation unit including a display unit 13 for displaying various operation screens on an operation panel on the front surface of the casing, and various operation keys and switches. 14 and a voice input unit 17 including a microphone or the like is provided.

(2) Electrical configuration
Next, FIG. 2 is a block diagram showing an electrical configuration of the route search apparatus 100. In this figure, the CPU 10 controls each part of the apparatus in accordance with an operation event supplied from the operation unit 14, and detailed description will be made later regarding the operation of characteristic processing according to the gist of the present invention. The ROM 11 stores various program data including a program for a main routine (see FIG. 3) and a route search process (see FIG. 4) which are executed by the CPU 10.

As shown in FIG. 3, the RAM 12 includes a work area WE for temporarily storing various register / flag data and a data area DE. The data area DE of the RAM 12 includes an address book ADR, a mail box MB, and a regional memory RM. Address data is stored in the address book ADR by the address book processing unit 20. The address data includes a telephone number, a postal code, and an address lot number. The mail processing unit 21 stores data of outgoing mail (sent mail) and incoming mail in the mailbox MB. A plurality of area names and their priorities are stored in the area memory RM in association with each other. The purpose of the contents of the regional memory RM will be described later.

  The display unit 13 displays various operation screens according to a display control signal supplied from the CPU 10. The operation unit 14 instructs to execute a power switch for powering on / off the apparatus power, an address registration switch for instructing activation of the address book processing unit 20, a mail transmission / reception switch for instructing activation of the mail processing unit 21, and route search processing. It has various switches such as a route search switch to generate an operation event corresponding to the type of switch to be operated. An operation event generated by the operation unit 14 is captured by the CPU 10.

  The route search database 15 stores the station name, the area name where the station is located, the route name, and the arrival and departure times of the train at each station. The landmark table 16 stores the names of landmarks for each region, and reads the corresponding region names using the landmark names as read addresses. The landmark mentioned here refers to the name of a famous building or commercial facility in the area, for example. In the area weighting process described later, when a name indicating a landmark exists in the transmitted / received mail data stored in the mailbox MB of the RAM 12, the area name having the landmark of the name is read from the landmark table 16. ing. That is, the landmark table 16 converts the name of the landmark into the name of the area where the landmark exists and outputs it.

  The audio input unit 17 includes a microphone disposed on the apparatus panel, and a conversion unit that amplifies an output signal of the microphone and then performs A / D conversion to generate audio data. The speech recognition unit 18 performs phoneme analysis on the speech data input from the speech input unit 17 under the control of the CPU 10, and recognizes the input speech as a word based on the analysis result. The communication unit 19 makes a wired connection (or wireless connection) to the Internet under the control of the CPU 10.

The address book processing unit 20 registers address data input in response to a user operation in the address book ADR (see FIG. 3) of the RAM 12, and searches and edits address data registered in the address book ADR. It should be noted that the address book processing unit 20 shown in this figure is not an independent hardware, but corresponds to a functional block embodied by the processing of the CPU 10.

The mail processing unit 20 is a functional block embodied by the processing of the CPU 10, similar to the address book processing unit 20. The mail processing unit 20 accesses a mail server on the Internet connected via the communication unit 19, and stores the mailbox MB ( The outgoing mail stored in FIG. 3) is sent to the mail server. On the other hand, if there is a received mail in the mail server, it is received and stored in the mailbox MB.

The route search processing unit 22 cooperates with the processing of the CPU 10 (route search processing to be described later), performs a route search with reference to the route search database 15 based on the boarding station name and the getting-off station name inputted by voice, and a search result. Is generated. The search result includes the name of each line connecting from the boarding station to the stop station, the arrival and departure times and the required time for each connecting line.

C. Operation Next, the operation of the first embodiment having the above-described configuration will be described with reference to FIGS. Hereinafter, as operations of the first embodiment, operations of a main routine executed by the CPU 10, route search processing called from the main routine, area weighting processing constituting the route search processing, voice input processing, and station name designation processing The operation will be described.

(1) Operation of Main Routine The operation of the main routine will be described with reference to FIG. When the route search device 100 is powered on in response to the power switch operation, the CPU 10 proceeds to step SA1 shown in FIG. 4 and initializes each part of the device. In step SA2 and subsequent steps, processing according to an operation event that occurs in response to a user operation is executed.

  That is, in step SA2, it is determined whether or not the address registration switch has been operated. If the address registration switch is not operated, the determination result is “NO”, and the process proceeds to step SA4, which will be described later. However, if the address registration switch is operated, the determination result is “YES”, and the process proceeds to step SA3. In step SA3, the address book processing unit 20 is activated according to the operation of the address registration switch.

As a result, the address book processing unit 20 registers the address data input in response to a user operation in the address book ADR (see FIG. 3) of the RAM 12, or searches and edits the address data registered in the address book ADR. Execute the address registration process. When the address registration process of the address book processing unit 20 is completed, the CPU 10 advances the process to step SA4.

  In step SA4, it is determined whether or not the mail transmission / reception switch has been operated. If the mail transmission / reception switch is not operated, the determination result is “NO”, and the process proceeds to step SA6 described later. However, if the mail transmission / reception switch is operated, the determination result is “YES”, and the process proceeds to step SA5. In step SA5, the mail processing unit 21 is activated according to the operation of the mail transmission / reception switch.

As a result, the mail processing unit 21 accesses the mail server on the Internet connected via the communication unit 19 and sends the outgoing mail stored in the mail box MB (see FIG. 3) of the RAM 12 to the mail server. If there is a received mail in the mail server, a mail transmission / reception process for receiving the received mail and storing it in the mailbox MB is executed. When the mail transmission / reception process of the mail processing unit 21 is completed, the CPU 10 advances the process to step SA6.

  In step SA6, it is determined whether or not the route search switch has been operated. If the route search switch is not operated, the determination result is “NO”, and the process proceeds to step SA8 described later. However, if the route search switch is operated, the determination result is “YES”, and the route search is performed via step SA7. Execute the process.

In the route search process, as will be described later, the user has some relationship with various data stored in the memory such as address data registered in the address book ADR of the RAM 12 and mail data stored in the mail box MB. Each time an area name that seems to be present is detected or specified, weighting is performed to increase the priority of the area name, and when the user inputs the station name (boarding station name / getting station name) by voice, the input voice data A station name that matches the word that has been subjected to phoneme analysis and recognized by speech is searched from the route search database 15.

When a plurality of station names are searched, among the station names, the station names existing in the area names that have some relationship with the user are weighted to increase the priority of the corresponding area names, and the weighted area names are prioritized. The names of multiple stations are rearranged in the order corresponding to the magnitude of the degree and displayed on the route search screen. By doing this, the station name that exists in the area name that has some relationship with the user becomes a higher selection candidate, so that it does not cause misrecognition or troublesome search, and the station name that is input by voice is recognized with high accuracy. It has come to be able to do.

When such a route search process is completed, the process proceeds to step SA8, and other processes such as adjusting the microphone sensitivity of the voice input unit 17 are executed. Thereafter, the above steps SA2 to SA8 are repeated until the apparatus power is turned off by operating the power switch.

(2) Route Search Processing Operation Next, route search processing operation will be described with reference to FIGS. When this process is executed via step SA7 (see FIG. 4) of the main routine described above, the CPU 10 executes the area weighting process of step SB1 shown in FIG. In the area weighting process, as will be described later, the weighting for increasing the priority of the area name that seems to have some relationship with the user by referring to the address data of the address book ADR provided in the RAM 12 or the contents of the mail data of the mail box MB. I do.

  Next, in step SB2, the route search processing unit 22 is instructed to start, and in the subsequent step SB3, the boarding station name and the getting-off station name are input to the route search screen (not shown) display unit 13 for instructing the search. . In step SB4, it is determined whether or not an input of the boarding station name is requested from the route search processing unit 22. When the input of the boarding station name is requested, the determination result here is “YES”, and the voice input process is executed via step SB5.

In the voice input process, as will be described later, the voice recognition unit 18 activates the voice input unit 17 and the voice recognition unit 18, and thereby the word recognized by the voice recognition unit 18 by performing phoneme analysis on the user's voice data captured by the voice input unit 17. And the station name that matches the extracted word is searched from the route search database 15.

  In step SB6, a station name designation process is executed. In the station name designation process, as will be described later, if there are multiple station names (boarding station names) that match the words obtained by speech recognition, the station names that exist in the area name that seems to have some relationship with the user among those station names Is weighted to increase the priority of the corresponding area name, and multiple station names (riding station names) are rearranged in the order corresponding to the priority of the weighted area name and displayed on the route search screen. Is selected by a user operation.

  Thus, when the boarding station is selected by the station name designation process in step SB6, the process proceeds to step SB7, and it is determined whether or not both stations (boarding station and getting-off station) have been selected. If only the boarding station is selected, the determination result is “NO”, and the process returns to step SB4 described above. In this case, since the boarding station name has already been selected, the determination result in step SB4 is “NO”, and the flow proceeds to step SB8.

  In step SB8, it is determined whether or not the route search processing unit 22 has requested input of the getting-off station name. If the input of the station name is requested, the determination result is “YES”, and the process proceeds to step SB9. In step SB9, the speech recognition unit 18 performs a phoneme analysis on the speech data of the user captured by the speech input unit 17 to extract a recognized word, and searches the route search database 15 for the name of the getting-off station that matches the extracted word. Perform voice input processing.

  Next, in step SB10, when there are a plurality of getting-off station names that match the word obtained by voice recognition, the priority of the corresponding area name is increased for the getting-off station name existing in the area name that seems to have some relationship with the user. A station name designation process is performed in which weighting is performed and a plurality of disembarkation station names are rearranged in the order corresponding to the priority of the weighted area names and displayed on the route search screen, and any of them is selected by a user operation. After execution, the process proceeds to step SB7.

  If the getting-off station is selected by the station name designation process in step SB10 and both the boarding station name and the getting-off station name are selected, the determination result in step SB7 is “YES”, and the flow advances to step SB11. In step SB11, it is determined which of the decision switch for determining the input of the selected boarding station name and the getting-off station name or the cancel switch for canceling the input has been operated. If the cancel switch is operated, the determination result is “NO”, and the process returns to step SB4 described above. When the cancel switch is operated, the route search processing unit 22 requests the CPU 10 to re-input the boarding station name.

  On the other hand, when the determination switch for determining the input of the boarding station name and the getting-off station name is operated, the judgment result at Step SB11 is “YES”, and the process proceeds to Step SB12 shown in FIG. In Step SB12, the boarding station name and the getting-off station name determined to be input are supplied to the route search processing unit 22 to instruct the start of the route search. Thereby, the route search processing unit 22 searches the route from the boarding station name to the getting-off station name with reference to the route search database 15.

Next, in step SB13, the process waits until the route search processing unit 22 completes the route search. When the route search in the route search processing unit 22 is completed, the determination result here is “YES”, the process proceeds to step SB14, and the search result is displayed on the route search screen. The search result includes the name of each line connecting from the boarding station to the stop station, the arrival and departure times and the required time for each connecting line.

Thereafter, when the process proceeds to step SB15, it is determined whether or not the return switch is instructed to perform the route search again. If the return switch is not operated, the determination result is “NO”, and the process proceeds to step SB17, which will be described later. However, if the return switch is operated to indicate the re-execution of the route search, the determination result in step SB15 is “ YES "and the process proceeds to Step SB16. In step SB16, after the search result displayed on the route search screen in step SB14 is cleared, the process is returned to step SB3 (see FIG. 5) and the route search is performed again from the beginning.

  Subsequently, in step SB17, it is determined whether or not the end switch is operated. If the end switch is not operated, the determination result is “NO”, and the process returns to step SB15. However, if the end switch is operated, the determination result is “YES” and the process proceeds to step SB18. The route search processing unit 22 is instructed to stop the processing. Then, the process proceeds to step SB19, the route search screen being displayed on the display unit 13 is cleared, and the process is finished.

  As described above, in the route search process, each time an area name that seems to have some relationship with the user is detected or specified by referring to the address data of the address book ADR or the mail data of the Mailbox MB, Gives weights that increase the priority of names. Next, when the user inputs the boarding station name by voice, a phoneme analysis is performed on the input voice data, and a boarding station name that matches the voice-recognized word is searched from the route search database 15.

And if there are multiple boarding station names that match the words obtained by speech recognition, weighting to increase the priority of the corresponding area name for the boarding station name that exists in the area name that seems to have some relationship with the user A plurality of boarding station names are rearranged in the order corresponding to the priority of the weighted area names and displayed on the route search screen. That is, the boarding station name existing in the area name that seems to have some relationship with the user is displayed on the screen as a higher selection candidate, and one of these selection candidates (station name) is selected by the user operation.

In the same manner as described above, the name of the getting-off station is displayed on the screen as an upper selection candidate in the area name that seems to have some relationship with the user, and one of these selection candidates (station name) is selected by the user operation. Is done. Search results including the name of each route connecting from the boarding station to the stop station, the departure and arrival times and the required time for each connecting route, according to the boarding station name and the getting-off station name selected in this way. Is displayed on the route search screen.

(3) Operation of Area Weighting Process Next, the operation of the area weighting process will be described with reference to FIGS. When this process is executed via step SB1 (see FIG. 5) of the route search process described above, the CPU 10 proceeds to step SC1 shown in FIG. 7, and the regional memory RM (FIG. 3) provided in the data area DE of the RAM 12 is displayed. Clear). Next, in step SC2, address data is extracted from the address book ADR provided in the data area DE of the RAM 12, and in the subsequent step SC3, the area name is detected from the extracted address data. Specifically, for example, an area name is detected by referring to a zip code in address data, or an area name is detected from an address lot number in address data. The area name refers to the name up to the municipality of the prefecture.

  In step SC4, it is determined whether or not the area name detected in step SC3 has been registered in the area memory RM. If the detected area name has been registered in the area memory RM, the determination result in step SC4 is “NO”, and the process proceeds to step SC5, where the priority corresponding to the registered area name is incremented and then described later. Proceed to step SC8.

  On the other hand, if the area name detected in step SC3 is not registered in the area memory RM, the determination result in step SC4 is “YES”, and the process proceeds to step SC6. In step SC6, the area name detected in step SC3 is newly registered in the area memory RM, and in step SC7, the priority corresponding to the newly registered area name is set to “1 (initial value)”.

  Thereafter, the process proceeds to step SC8, and it is determined whether or not the area names have been detected for all address data in the address book ADR provided in the data area DE of the RAM 12. If the area name has not been detected for all address data, the determination result is “NO”, and the process returns to step SC2 to repeat the above-described process. When the detection of the area name is completed for all address data, the determination result in step SC8 is “YES”, and the process proceeds to step SC9 shown in FIG.

  In step SC9, mail data is extracted from the mail box MB (see FIG. 3) provided in the data area DE of the RAM 12. The mail data includes sent mail data and received mail data. Next, in step SC10, it is determined whether or not the contents specifying the region are included in the mail data extracted from the mail box MB. Specifically, a name (proper noun) representing a landmark is searched from the mail text. If the contents specifying the area are not included in the mail data, that is, if the name representing the landmark cannot be searched from the mail text, the determination result is “NO”, and the process proceeds to Step SC16 described later.

  On the other hand, if the name representing the landmark can be retrieved from the mail text, the determination result in step SC10 is “YES”, and the flow advances to step SC11. In step SC11, the landmark table 16 is referred to and the name of the searched landmark is converted into the area name where the landmark is located. When the area name is specified from the landmark name included in the mail data in this way, the process proceeds to step SC12, and it is determined whether or not the specified area name has been registered in the area memory RM.

If the specified area name has already been registered in the area memory RM, the determination result in step SC12 is “NO”, and the process proceeds to step SC13 to increment the priority corresponding to the registered area name, which will be described later. The process proceeds to step SC16. On the other hand, if the specified area name is not registered in the area memory RM, the determination result in step SC12 is “YES”, and the flow advances to step SC14. In step SC14, the area name identified in step SC11 is newly registered in the area memory RM, and in the subsequent step SC15, the priority corresponding to the newly registered area name is set to "1 (initial value)".

  Thereafter, the process proceeds to step SC16, and it is determined whether or not the search has been completed for all the mail data in the mail box MB provided in the data area DE of the RAM 12. If the search has not been completed, the determination result is “NO”, the process returns to step SC9, and the above-described process is repeated. When all the mail data has been searched, the determination result in step SC16 is “YES”, and this process is completed.

  As described above, in the area weighting process, each time the area name is detected from the address data registered in the address book ADR of the RAM 12 using the area memory RM of the RAM 12 in which the area name and its priority are associated and stored. The priority corresponding to the detected area name is incremented and weighted, and every time the area name is specified from the mail data stored in the mail box MB of the RAM 12, the priority corresponding to the specified area name is incremented. And weight. That is, each time an area name that seems to have some relationship with the user is detected or specified by referring to the address data of the address book ADR or the mail data of the mail box MB, weighting for increasing the priority of the area name is performed. Done.

  In other words, the station name that is usually entered for route search is a region where the user has some involvement in the past (for example, there are acquaintances and friends who have been in contact, with a few exceptions) , There are companies with business relations). Therefore, in the area weighting process, in view of these points, weighting is performed to increase the priority of the area name that seems to have some relationship with the user by referring to the address data of the address book ADR and the mail data of the Mailbox MB. That is why.

(4) Operation of Voice Input Process Next, the operation of the voice input process will be described with reference to FIG. When this process is executed through step SB5 (see FIG. 5) of the route search process described above, the CPU 10 proceeds to step SD1 shown in FIG. 9, and instructs the voice input unit 17 and the voice recognition unit 18 to start. . Subsequently, in step SD2, the process waits until a voice is input. Then, it is assumed that the user inputs the boarding station name by voice. Then, the voice recognition unit 18 performs phoneme analysis on the voice data captured by the voice input unit 17, recognizes the input voice based on the analysis result, and notifies the CPU 10 of the recognition completion.

  When the voice recognition unit 18 notifies the CPU 10 of the completion of recognition, the determination result in step SD2 is “YES”, and the process proceeds to step SD3, where the words recognized by the voice recognition unit 18 are extracted. Next, in step SD4, a station name that matches the extracted word is searched from the route search database 15, and the process is terminated.

  As described above, in the voice input process, the voice input unit 17 and the voice recognition unit 18 are activated, and thereby the words recognized by the voice recognition unit 18 performing phoneme analysis on the voice data of the user captured by the voice input unit 17 are recognized. The station name that matches the extracted word is retrieved from the route search database 15.

(5) Operation of station name designation process Next, the operation of the station name designation process will be described with reference to FIG. When this process is executed via either step SB6 (see FIG. 5) or step SB10 (see FIG. 5) of the route search process described above, the CPU 10 proceeds to step SE1 shown in FIG. It is determined whether or not a plurality of station names matching the obtained word have been searched from the route search database 15. If there is only one station name searched from the route search database 15, the determination result is “NO”, the process proceeds to step SE2, and the searched station name is displayed in the boarding station input area on the route search screen. indicate.

  On the other hand, if there are a plurality of station names searched from the route search database 15, the determination result of step SE1 is “YES”, and the process proceeds to step SE3. In step SE3, one of the searched station names is designated. Subsequently, in step SE4, the area name where the designated station name exists is extracted from the route search database 15. In step SE5, it is determined whether or not the region name extracted in step SE4 is registered in the region memory RM (see FIG. 3) provided in the data area DE of the RAM 12.

If the extracted area name is registered in the area memory RM, the determination result in step SE5 is “YES”, and the flow advances to step SE6. In step SE6, weighting is performed to increase the priority corresponding to the region name (extracted region name) registered in the region memory RM, and then the process proceeds to step SE7. Note that the weighting for increasing the priority refers to a process of adding a predetermined weighting value to the priority, for example.

On the other hand, if the extracted region name is not registered in the region memory RM, the determination result in step SE5 is “NO”. In this case, the process proceeds to step SE7 without performing weighting. In step SE7, it is determined whether or not the search has been completed for all of the plurality of station names searched from the route search database 15. If the search has not been completed, the determination result is “NO”, and the processing from step SE3 onward is repeated.

When the search is completed for all of the plurality of station names searched from the route search database 15, the determination result at Step SE7 is “YES”, and the process proceeds to Step SE8. In step SE8, the contents of the weighted area memory RM are referred to, and a plurality of station names searched from the route search database 15 are rearranged in the order of priority of the corresponding area names in the route search screen. To display. Therefore, the station name registered in the area memory RM and existing in the area name that seems to have some relationship with the user is displayed on the screen as an upper selection candidate. Next, in step SE9, the user selects a desired station name from among a plurality of station names displayed on the route search screen, and the process is completed.

  Thus, in the station name designation process, if there are multiple station names that match the word obtained by speech recognition, the priority of the corresponding area name is set for the station name that exists in the area name that seems to have some relationship with the user. The weight is increased, and a plurality of station names are rearranged in the order corresponding to the priority of the weighted area names and displayed on the route search screen. That is, the station name existing in the area name that seems to have some relationship with the user is displayed on the screen as a higher selection candidate, and any one of these selection candidates (station names) is selected by the user operation.

  As described above, in the first embodiment, there is some relationship to the user from various data stored in memory by the user, such as address data registered in the address book ADR and mail data stored in the mail box MB. Every time an area name that seems to be present is detected or specified, weighting is performed to increase the priority of the area name, and when the user inputs a station name (boarding station name / getting station name) by voice, phoneme analysis is performed on the input voice data. The station name that matches the speech-recognized word is searched from the route search database 15.

When multiple station names are searched, among those station names, for the station names that exist in the area name that seems to have some relationship with the user, weighting is performed to increase the priority of the corresponding area name, and the weighted area By rearranging multiple boarding station names in the order corresponding to the priority of the names and displaying them on the route search screen, the station names existing in the area name that seems to have some relationship with the user are selected as the top selection candidates. As a result, erroneous recognition and troublesome search are eliminated, and the station name input by voice can be recognized with high accuracy.

[Second Embodiment]
Next, a second embodiment will be described with reference to FIGS. FIG. 11 is a block diagram showing an electrical configuration of the route search device 100 according to the second embodiment. In this figure, the same number is attached | subjected to the same component as 1st Embodiment shown in FIG. 2, and the description is abbreviate | omitted.

The second embodiment is different from the first embodiment described above in that a GPS receiver 23 that receives a GPS satellite signal and measures the current position under the control of the CPU 10 and a GPS provided in the data area DE of the RAM 12. And a memory. In the GPS memory, as will be described later, a region name including a place (position) with a visit history, and a priority level corresponding to each region name and increased according to the number of visits are registered.

  Next, the operation of the second embodiment will be described. Hereinafter, operations different from those of the first embodiment, that is, operations of the main routine of the second embodiment, the area weighting process of the second embodiment, and the GPS process will be described. In addition, since each operation | movement of a route search process except a region weighting process, a voice input process, and a station name designation | designated process is the same as the above-mentioned 1st Embodiment, it abbreviate | omits about the description.

(1) Operation of Main Routine of Second Embodiment FIG. 12 is a flowchart showing the operation of the main routine of the second embodiment. As in the first embodiment described above, when the route search device 100 is powered on in response to the power switch operation, the CPU 10 proceeds to step SF1 shown in FIG. 12 and initializes each part of the device.

  Subsequently, in step SF2, it is determined whether or not the address registration switch has been operated. If the address registration switch is not operated, the determination result is “NO”, and the process proceeds to Step SF4 described later. However, if the address registration switch is operated, the determination result is “YES”, and the process proceeds to Step SF3. In step SF3, the address book processing unit 20 is activated according to the operation of the address registration switch.

As a result, the address book processing unit 20 registers the address data input in response to a user operation in the address book ADR (see FIG. 3) of the RAM 12, or searches and edits the address data registered in the address book ADR. Execute the address registration process. When the address registration process of the address book processing unit 20 is completed, the CPU 10 advances the process to step SF4.

  In step SF4, it is determined whether or not the mail transmission / reception switch has been operated. If the mail transmission / reception switch is not operated, the determination result is “NO”, and the process proceeds to step SF6 described later. However, if the mail transmission / reception switch is operated, the determination result is “YES”, and the process proceeds to step SF5. In step SF5, the mail processing unit 21 is activated according to the operation of the mail transmission / reception switch.

As a result, the mail processing unit 21 accesses the mail server on the Internet connected via the communication unit 19 and sends the outgoing mail stored in the mail box MB (see FIG. 3) of the RAM 12 to the mail server. If there is a received mail in the mail server, a mail transmission / reception process for receiving the received mail and storing it in the mailbox MB is executed. When the mail transmission / reception process of the mail processing unit 21 is completed, the CPU 10 advances the process to step SF6.

  In step SF6, it is determined whether or not the route search switch has been operated. If the route search switch is not operated, the determination result is “NO”, and the process proceeds to step SF8, which will be described later. However, if the route search switch is operated, the determination result is “YES”, and the second result is obtained via step SF7. Route search processing according to the embodiment is executed.

In the route search processing according to the second embodiment, there is some relationship to the user from various data stored in the memory by the user, such as address data registered in the address book ADR of the RAM 12 and mail data stored in the mail box MB. Every time an area name that seems to be present is detected or specified, weighting is performed to increase the priority of the area name, and then the priority of the area name that has been visited in the past is increased based on the contents of the GPS memory. . Then, when the user inputs the station name (boarding station name / getting off station name) by voice, the input voice data is subjected to phoneme analysis and the station name matching the voice recognized word is searched from the route search database 15.

When a plurality of station names are searched, weights that increase the priority of the corresponding area names are given to the station names that have some relationship with the user and that exist in the area names that have been visited. By arranging multiple boarding station names in the order corresponding to the priority of the area name and displaying it on the route search screen, the station name that has some relationship with the user and exists in the area name that has a history of visiting is higher. Therefore, it is possible to recognize the station name input by voice with high accuracy, without causing erroneous recognition and troublesome search.

When such route search processing is completed, GPS processing is executed via step SF8. In the GPS processing, as will be described later, an area name including the current position measured every certain period is extracted. If the extracted area name is not registered in the GPS memory, it is newly registered in the GPS memory and newly registered. The priority corresponding to the area name is set to “1”. On the other hand, if the extracted area name is already registered in the GPS memory, the priority corresponding to the area name is incremented to raise the priority. By doing so, locations (positions) that have been visited in the past, that is, names of regions that have been visited, and priorities that correspond to these region names and that can be increased according to the number of visits are registered in the GPS memory. If there is an area name that has not been revisited for a predetermined period, it is deleted from the GPS memory.

When the GPS process is completed, the process proceeds to step SF9, and other processes such as adjusting the microphone sensitivity of the voice input unit 17 are executed. Thereafter, steps SF2 to SF9 are repeated until the apparatus power is turned off by the power switch operation.

(2) Operation of Area Weighting Process According to Second Embodiment Next, the operation of the area weighting process according to the second embodiment will be described with reference to FIGS. As in the first embodiment described above, when this process is executed via step SB1 (see FIG. 5) of the route search process, the CPU 10 proceeds to step SG1 shown in FIG. 13 and is provided in the data area DE of the RAM 12. The area memory RM (see FIG. 3) to be used is cleared. Next, in step SG2, address data is extracted from the address book ADR provided in the data area DE of the RAM 12, and in the subsequent step SG3, the area name is detected from the extracted address data. Specifically, for example, an area name is detected by referring to a zip code in address data, or an area name is detected from an address lot number in address data. The area name refers to the name up to the municipality of the prefecture.

  In step SG4, it is determined whether or not the area name detected in step SG3 has been registered in the area memory RM. If the detected area name is already registered in the area memory RM, the determination result in step SG4 is “NO”, and the process proceeds to step SG5. After the priority corresponding to the registered area name is incremented, step SG8 is performed. Proceed to

  On the other hand, if the area name detected in step SG3 is not registered in the area memory RM, the determination result in step SG4 is “YES”, and the process proceeds to step SG6. In step SG6, the area name detected in step SG3 is newly registered in the area memory RM, and in the subsequent step SG7, the priority corresponding to the newly registered area name is set to “1 (initial value)”.

  Thereafter, the process proceeds to step SG8, where it is determined whether or not the area name has been detected for all address data in the address book ADR provided in the data area DE of the RAM 12. If the area name has not been detected for all address data, the determination result is “NO”, and the process returns to step SG2 to repeat the above-described process. Then, when the region names have been detected for all the address data, the determination result in step SG8 is “YES”, and the process proceeds to step SG9 shown in FIG.

  In step SG9, mail data is extracted from the mail box MB (see FIG. 3) provided in the data area DE of the RAM 12. The mail data includes sent mail data and received mail data. Next, in step SG10, it is determined whether or not the contents specifying the area are contained in the mail data extracted from the mail box MB. Specifically, a name representing a landmark is searched from the mail text. If the content specifying the area is not included in the mail data, that is, if the name representing the landmark cannot be searched from the mail text, the determination result is “NO”, and the process proceeds to Step SG16.

  On the other hand, if the name representing the landmark can be retrieved from the mail text, the determination result in step SG10 is “YES”, and the flow proceeds to step SG11. In step SG11, the landmark table 16 is referred to, and the name of the searched landmark is converted into the area name where the landmark is located. Thus, when the area name is specified from the landmark name included in the mail data, the process proceeds to step SG12, and it is determined whether or not the specified area name has been registered in the area memory RM.

If the specified region name has already been registered in the region memory RM, the determination result in step SG12 is “NO”, and the process proceeds to step SG13. After the priority corresponding to the registered region name is incremented, step Proceed to SG16. On the other hand, if the specified region name is not registered in the region memory RM, the determination result in step SG12 is “YES”, and the process proceeds to step SG14. In step SG14, the area name specified in step SG11 is newly registered in the area memory RM, and in the subsequent step SG15, the priority corresponding to the newly registered area name is set to “1 (initial value)”.

  Thereafter, the process proceeds to step SG16, where it is determined whether or not the search has been completed for all mail data in the mail box MB provided in the data area DE of the RAM 12. If the search has not been completed, the determination result is “NO”, and the process returns to step SG9 to repeat the above-described process. When all the mail data has been searched, the determination result in step SG16 is “YES”, and the process proceeds to step SG17. In step SG17, weighting is performed to change the priority of the area name registered in the area memory RM in accordance with the contents of the GPS memory, and the process ends.

  That is, the GPS memory stores the names of areas that have been visited by GPS processing described later. Therefore, the weighting for changing the priority of the region name registered in the region memory RM according to the contents of the GPS memory is the priority of each region name obtained from the address data of the address book ADR or the mail data of the Mailbox MB. Among them, it refers to increasing the priority of area names that have been in the past.

  As described above, the area weighting process according to the second embodiment seems to have some relationship with the user by referring to the address data of the address book ADR and the mail data of the mail box MB, as in the first embodiment. Each time a region name is detected or specified, weighting is performed to increase the priority of the region name, and then the priority of the region name having a visit history is increased based on the contents of the GPS memory.

(3) Operation of GPS processing Next, the operation of GPS processing will be described with reference to FIG. When this process is executed via step SF8 (see FIG. 12) of the main routine according to the second embodiment, the CPU 10 proceeds to step SH1 shown in FIG. 15 and determines whether or not it is the activation timing of the GPS receiver 23. Judging. If it is not under the start timing, the determination result is “NO”, and this process is finished without doing anything. However, when the start timing is reached (for example, every 12 hours), the determination result in step SH1 is “YES”. The process proceeds to step SH2.

  In step SH2, the activation is instructed to the GPS receiver 23, and in the subsequent step SH3, the apparatus waits until the activated GPS receiver 23 receives a GPS satellite signal and generates position information. When the GPS receiving unit 23 generates position information, the determination result in step SH3 is “YES”, and the process proceeds to step SH4 to acquire the position information generated by the GPS receiving unit 23, that is, the current position. Next, in step SH5, an area name including the acquired current position is extracted.

  Subsequently, in step SH6, it is determined whether or not the area name extracted in step SH5 has been registered in the GPS memory. If the extracted region name is not registered in the GPS memory, the determination result is “NO”, and the process proceeds to step SH7, where the region name extracted in step SH5 is newly registered in the GPS memory, and the newly registered region name After the priority corresponding to is set to “1 (initial value)”, the process proceeds to step SH9.

  On the other hand, if the extracted area name is already registered in the GPS memory, the determination result in step SH6 is “YES”, and the process proceeds to step SH8, in which the priority corresponding to the registered area name is incremented and the priority order is increased. After increasing the value, the process proceeds to step SH9. In step SH9, it is determined whether or not there is an area name that is not extracted for a predetermined period or more among the area names registered in the GPS memory.

That is, since the area names with a history of visit are registered in the GPS memory, in this step SH9, it is determined whether or not there is an area name that has not been visited for a predetermined period among the area names with a history of visit. If there is no area name that has not been visited for a predetermined period of time, the determination result here is “NO”, and this process ends. If there is an area name that has not been visited for a predetermined period of time, the determination result is “YES”, and step SH10 Proceed to, the corresponding area name is deleted from the GPS memory, and the process is terminated.

  As described above, in the GPS processing, the region name including the current position measured every certain period is extracted, and if the extracted region name is not registered in the GPS memory, the region is newly registered in the GPS memory and the newly registered region The priority corresponding to the name is set to “1”. On the other hand, if the extracted area name is already registered in the GPS memory, the priority corresponding to the area name is incremented to raise the priority. By doing so, the names of regions having a history of visits and the priorities corresponding to these region names and increased according to the number of visits are registered in the GPS memory. If there is an area name that has not been visited for a predetermined period, the area name is deleted from the GPS memory.

  As described above, in the second embodiment, there is some relationship to the user from various data stored in the memory by the user, such as address data registered in the address book ADR and mail data stored in the mail box MB. Each time an area name that is considered to be detected or specified, weighting is performed to increase the priority of the area name, and then the priority of the area name with a history of visit is increased based on the contents of the GPS memory. That is, the priority of the area name which has a certain relationship with the user and who has visited is increased.

Then, when the user inputs the station name (boarding station name / getting off station name) by voice, the input voice data is subjected to phoneme analysis to search the route search database 15 for a station name that matches the voice recognized word, and a plurality of station names are searched. Is searched, station names that have a certain relationship with the user among those station names and that exist in the area names that have been visited are weighted to increase the priority of the corresponding area names, and the weighted area names A plurality of boarding station names are rearranged in the order corresponding to the priority level and displayed on the route search screen. As a result, the station name that has some relationship with the user and that exists in the area name that has been visited becomes a high-order selection candidate, so that it does not cause misrecognition or troublesome search, and is input by voice. The station name can be recognized with high accuracy.

  In the first and second embodiments described above, an area name that seems to have some relationship with the user is detected or detected based on address data registered in the address book ADR and mail data stored in the mail box MB. However, the present invention is not limited to this. For example, when a web browser function is provided, a web page browsed by the function is saved, and an area name that seems to have some relationship with the user from the saved web page is selected. It may be detected or specified. It is also possible to save the map data obtained from the map site and extract an area name that seems to have some relationship with the user from there.

1 is an external view showing a schematic external view of a route search device 100 according to a first embodiment. It is a block diagram which shows the electric constitution of the route search apparatus 100 by 1st Embodiment. 3 is a memory map showing a memory configuration of a RAM 12; It is a flowchart which shows operation | movement of the main routine of 1st Embodiment. It is a flowchart which shows operation | movement of the route search process of 1st Embodiment. It is a flowchart which shows operation | movement of the route search process of 1st Embodiment. It is a flowchart which shows the operation | movement of the area weighting process of 1st Embodiment. It is a flowchart which shows the operation | movement of the area weighting process of 1st Embodiment. It is a flowchart which shows the operation | movement of the audio | voice input process of 1st Embodiment. It is a flowchart which shows operation | movement of the station name designation | designated process of 1st Embodiment. It is a block diagram which shows the electric constitution of the route search apparatus 100 by 2nd Embodiment. It is a flowchart which shows operation | movement of the main routine of 2nd Embodiment. It is a flowchart which shows the operation | movement of the area weighting process of 2nd Embodiment. It is a flowchart which shows the operation | movement of the area weighting process of 2nd Embodiment. It is a flowchart which shows operation | movement of the GPS process of 2nd Embodiment.

Explanation of symbols

100 Route Search Device 10 CPU
11 ROM
12 RAM
DESCRIPTION OF SYMBOLS 13 Display part 14 Operation part 15 Route search database 16 Landmark table part 17 Voice input part 18 Voice recognition part 19 Communication part 20 Address book process part 21 Mail process part 22 Route search process part 23 GPS receiving part

Claims (3)

User data storage means for storing a plurality of user data containing contents capable of deriving a name related to the user;
Search means for searching for a name that matches the result of recognizing the input speech;
When the search means searches for a plurality of candidate names, based on the contents of the plurality of user data stored in the user data storage means, weighting means for obtaining weighted priorities of the searched candidate names;
Name display means for rearranging and displaying a plurality of candidate names searched by the search means based on the priority level of the candidate names weighted by the weighting means;
Comprising
The user data storage means includes a first storage means for storing a plurality of first type user data including address data registered in response to a user operation, and a second storage means for storing a plurality of second type user data. Including,
Said weighting means, prior to retrieval by the retrieval means, based on the address data included in the plurality of first type user data stored in said first storage means, obtains the priority of each of a plurality of names, further , the priority for each name, the second storage means by changing based on the presence or absence of the name included in the stored plurality of second type user data in advance stored Sakido Yu of each name In addition, when a plurality of candidate names are searched by the search means, the weight of each searched candidate name is determined based on the priority of each name stored in advance corresponding to each candidate name. Ask,
A search device characterized by that. The second type user data is sent and received mail data.
The search device according to claim 1. A program for controlling a computer of a search device,
The computer,
User data storage means for storing a plurality of user data containing contents capable of deriving a name related to the user;
Search means for searching for a name that matches the result of recognizing the input speech,
When the search means searches for a plurality of candidate names, based on the contents of the plurality of user data stored in the user data storage means, weighting means for obtaining weighted priorities of the searched candidate names;
Name display means for rearranging and displaying a plurality of candidate names searched by the search means based on a priority level of the candidate names weighted by the weighting means;
Function as
The user data storage means includes a first storage means for storing a plurality of first type user data including address data registered in response to a user operation, and a second storage means for storing a plurality of second type user data. Including,
Said weighting means, prior to retrieval by the retrieval means, based on the address data included in the plurality of first type user data stored in said first storage means, obtains the priority of each of a plurality of names, further , the priority for each name, the second storage means by changing based on the presence or absence of the name included in the stored plurality of second type user data in advance stored Sakido Yu of each name In addition, when a plurality of candidate names are searched by the search means, the weight of each searched candidate name is determined based on the priority of each name stored in advance corresponding to each candidate name. Ask,
A computer-readable program designed to function as follows.

Images