JP4689401B2 - Information retrieval device - Google Patents

Description

  The present invention relates to an information retrieval apparatus that retrieves information based on conditions input by voice, a keyboard, a touch panel, or the like and presents the retrieved information to a user.

  2. Description of the Related Art Conventionally, for example, an information retrieval device that retrieves information based on conditions input by a user using voice, a keyboard, a touch panel, or the like and presents the retrieved information, such as a navigation device, is known. In such an information search device, the searched information is visually displayed as a list on the screen or presented audibly by voice, for example. At this time, if there is a large amount of information that matches the input condition, it is difficult to present the retrieved information so that the user can easily recognize it. For this reason, an information search apparatus has been proposed that can narrow down the searched information when a large number of information matching the conditions is searched (see, for example, Patent Document 1).

  The navigation device as an information search device of Patent Literature 1 searches for registered locations (parks, stations, hotels, etc.) under the input conditions, displays the searched registered locations on the screen, and displays the registered locations displayed. It presents information such as a map around the registered location selected by the user. The navigation device includes a search unit that searches for an input character as a search condition and searches for a registration site that includes the character in the name, and a registration site that satisfies a more detailed filtering condition from the registration site searched by the search unit. And a narrowing search means for searching. Then, the navigation device searches for registered locations that include the input characters in the name, and when the number of registered locations searched is large, the searched registered locations are not displayed on the screen, and the genre or capital is displayed. A plurality of predetermined narrowing-down conditions such as prefecture names are displayed on the screen so that the user can arbitrarily specify them. Further, when the narrowing condition is selectively designated by the user, the navigation device narrows down and searches the registered locations that satisfy the designated narrowing condition, and displays the searched registered locations on the screen.

However, in the navigation device of Patent Document 1, when the number of registered locations searched is large, a plurality of predetermined narrowing conditions are presented so that the user can selectively specify them. Is. That is, in the navigation device of Patent Document 1, a plurality of predetermined narrowing conditions are simply listed when presenting narrowing conditions. Then, it is left to the user which narrowing condition to specify among the multiple types of narrowing conditions presented. Therefore, depending on the specified filtering conditions, the number of registered locations searched can be displayed on the screen even if the search is narrowed down (the number that can be easily recognized by the user without an excessive number of displayed items). ) May not be reduced. Also, depending on the specified narrowing conditions, there may be no registered place that satisfies the narrowing conditions. For this reason, the search location must not be presented properly, or the user must try to search many times by changing the filtering conditions so that the number of search locations that can be searched is appropriate. There was an inconvenience of not becoming.
JP 2002-107171 A

  The present invention eliminates such inconvenience, optimizes the search conditions for narrowing down the search target, and appropriately presenting information while minimizing the number of searches and the time required for the search. It is an object of the present invention to provide an information retrieval apparatus that can be used.

In order to achieve such an object, the information search apparatus of the present invention includes a storage unit for storing a search target, and an input unit for inputting a search condition for searching the search target stored in the storage unit from a user. A search object extraction unit for extracting a search object corresponding to the search condition input to the input unit, and a first number calculation for calculating a first number that is the number of search objects extracted by the search object extraction unit Means for narrowing search conditions that can be input to the input means for narrowing down the search target extracted by the search target extraction means, and narrowing condition candidate setting means For each search condition candidate set by the setting means, a second number that is the number of search targets corresponding to the search condition candidates among the search targets extracted by the search target extraction means is calculated. And 2 number calculating means is a means for determining the narrowing-down condition for notifying the candidate of the previous SL refinement condition to the user, of candidates for the narrowing-down condition, the ratio of the second count to the first count Is determined by the narrowing-down condition determining means for determining a narrowing-down condition candidate existing in the predetermined range set according to the first number as the narrowing-down condition to notify the user, and the narrowing-down condition determining means Informing the user of the narrowed down conditions, and the predetermined range is set according to the first number so that the upper limit of the range is lower as the first number is larger It is the range which was made .

According to the information search device of the present invention, the narrowing-down condition candidate setting unit is configured to search for a narrowing-down condition that can be input to the input unit for narrowing down the search target extracted by the search target extracting unit. Candidates are set, and the narrowing-down condition determining means determines narrowing-down conditions to be notified to the user from the narrowing-down condition candidates. At this time, the first number calculating means calculates a first number that is the number of search objects extracted by the search object extracting means, and the second number calculating means is a search object extracted by the search object extracting means. The second number, which is the number of search targets corresponding to the narrowing condition candidate, is calculated .
Its to the refinement condition determining means, based on the first count and the second count, to determine the narrowing-down condition for notifying the user from the candidate of the narrowing-down condition. More specifically, the narrowing-down condition determining means has a predetermined range in which the ratio of the second number to the first number of candidates for the narrowing-down condition is set according to the first number (the first number A narrowing condition candidate existing in a range set such that the upper limit of the range becomes lower as the number of cases increases, is determined as a narrowing condition to be notified to the user. For this reason, the narrowing-down condition notified to the user is determined such that the larger the first number is, the more the second number is narrowed. The determined narrowing condition is notified to the user by the notification means.
Therefore, the refinement condition determining means, narrow the search target, as can the number that can present information appropriately (the number that the user can easily recognize), the optimal narrowing conditions, grain It is possible to determine from the candidates for the narrowing conditions, and the determined narrowing conditions can be notified to the user by the notification means . Then, thus informing been narrowing conditions because it is specified by the user, even if the search target extracted by the search target extraction means have a large number of search target is narrowed down efficiently. Thereby, the information search device of the present invention can present information appropriately while minimizing the number of times of search and the time required for the search.

  Further, in the information search device of the present invention, the notifying means includes the second condition of the narrowing condition calculated by the second number calculating means together with the narrowing condition determined by the narrowing condition determining means. It is preferable to provide means for notifying the user of the number of cases.

  According to this, together with the narrowing condition, the second number of the narrowing condition (the number of search targets corresponding to the narrowing condition among the search targets extracted by the search target extracting unit) is used. The person is informed. Therefore, the user can know in advance how much the search target is narrowed down by the narrowing condition when specifying the narrowing condition, so that the narrowing condition can be specified more easily and information can be searched efficiently. be able to.

  The information search device of the present invention is mounted on a moving body such as a vehicle as a navigation device, and is used for searching for information by a passenger of the moving body. At this time, since there is usually an operation of a handle or the like during driving, it is desirable that operations such as input of search conditions can be easily performed. Therefore, when the information search device of the present invention is mounted on a mobile body, it is preferable that the input means input a search condition by the voice of the user who is an occupant of the mobile body.

  According to this, since the user can input the search condition by voice, for example, the search condition can be input without using a hand. Therefore, the user can easily input the search condition even during driving.

Furthermore, during driving, for example, in the case of a vehicle, the degree to which the driver must concentrate on driving differs depending on the driving load of the driver due to the traveling state of the vehicle. Therefore, when the information search device of the present invention is an information search device mounted on a moving body, driving load detection means for detecting the driving load of the user who is a driver among the passengers of the moving body is provided. Further , the predetermined range has an upper limit of the range as the operating load detected by the operating load detection means is higher in addition to the lower limit of the range being lower as the first number is larger. Is preferably in a range set according to the first number and the operating load .

According to this, the driving load of the user who is the driver of the moving body is detected by the driving load detection means. Here, for example, in the case of a vehicle, the driving load can be detected from vehicle speed, steering operation, braking operation, palm sweating, and the like. The narrowing-down condition determining means determines narrowing-down conditions to be notified to the user based on the driving load in addition to the first number and the second number. More specifically, the narrowing-down condition determining means has a predetermined range in which the ratio of the second number of cases to the first number of candidates for the narrowing-down conditions is set according to the first number and the operating load ( In addition to the upper limit of the range being lower as the first number is larger, the range is set such that the upper limit of the range is lower as the driving load detected by the driving load detecting means is higher. ) Is determined as a narrowing condition to be notified to the user.
Thereby, since the optimum narrowing-down condition considering the driving load of the user who is the driver is determined from the narrowing-down condition candidates and notified to the user, the user can Information can be searched efficiently. More specifically, when the driving load of the user who is the driver is high and it is necessary to concentrate on driving, a narrowing condition that can narrow down the search target is determined from the narrowing condition candidates and used. The person is informed. As a result, the number of searches and the time required for the search are further shortened, and the user can efficiently search for information even during driving.

  An embodiment of the present invention will be described with reference to the accompanying drawings. FIG. 1 is a functional block diagram of the information search apparatus according to the present embodiment, and FIGS. 2 and 3 are flowcharts of information search processing in the information search apparatus of FIG. 4 shows an example of voice guidance in the information search apparatus of FIG. 1, and FIG. 5 shows an example of a screen in the information search apparatus of FIG.

  Referring to FIG. 1, an information retrieval apparatus 1 is an electronic unit composed of a microcomputer or the like, and as its processing functions, a database 2 for storing a retrieval object and a retrieval condition for retrieving the retrieval object are input. Voice input I / F (interface) 3, search object extraction means 4 for extracting search objects corresponding to the search conditions, and first number calculation means for calculating a first number which is the number of extracted search objects 5, a narrowing condition candidate setting means 6 for setting a narrowing condition candidate for narrowing down the extracted search target, and the narrowing of the extracted search targets for each narrowing condition candidate A second number calculation means 7 for calculating a second number that is the number of search targets corresponding to the inclusion condition candidates is provided on the vehicle 12. Furthermore, the information retrieval apparatus 1 uses the driving load detection means 8 for detecting the driving load of the driver of the vehicle 12, the first number, the second number, and the driving load, from the narrowing condition candidates. The narrowing condition determining means 9 for determining the narrowing condition, the audio output I / F (interface) 10 for notifying the narrowing condition, and the display 11 are provided.

  The database 2 (storage means of the present invention) is a storage medium such as a CD-ROM, DVD, or HDD in which data to be searched is recorded. The search target recorded in the database 2 is a destination such as a restaurant, a station, a hotel, or a park, for example. In the database 2, data used as search conditions such as the position of the search target associated with the map data, the name of the search target, and the genre are recorded as the search target data. Further, the database 2 stores detailed information on the search target and information necessary for guiding the travel route to the search target as the search target data.

  In the voice input I / F 3 (input means of the present invention), a search condition for searching for a search target stored in the database 2 is inputted by voice from the driver of the vehicle 12. And about the driver | operator's speech input into voice input I / F3, voice recognition processes, such as a syntax analysis, are performed and search conditions are acquired.

  The search target extraction unit 4 extracts a search target corresponding to the search condition input to the voice input I / F 3 from the search targets stored in the database 2.

  The first number calculation means 5 calculates the first number that is the number of search objects extracted by the search object extraction means 4.

  The narrow-down condition candidate setting means 6 sets narrow-down condition candidates for narrowing down the search target extracted by the search target extraction means 4. The narrowing-down conditions are those that allow voice input from the driver to the voice input I / F 3 as search conditions.

  The second number calculation means 7 corresponds to the narrowing condition candidates among the search targets extracted by the search target extraction means 4 for each narrowing condition candidate set by the narrowing condition candidate setting means 6. A second number, which is the number of search targets, is calculated.

  The driving load detection means 8 detects the driving load of the driver based on the vehicle speed, steering operation, braking operation, palm sweating, and the like detected by a sensor or the like. The driving load of the driver is a value indicating the degree to which the driver concentrates on driving and the degree to which the driver should concentrate on driving. At this time, for example, the higher the vehicle speed, the greater the steering operation, the greater the brake operation, and the greater the perspiration of the palm, the higher the driving load.

  The narrowing-down condition determination means 9 includes the first number of cases calculated by the first number of cases calculation means 5, the second number of cases calculated by the second number of cases calculation means 7, and the driver's driving detected by the driving load detection means 8. Based on the load, the narrowing condition to be notified to the driver is determined from the narrowing condition candidates set by the narrowing condition candidate setting means 6. At this time, the narrowing-down condition determining means 9 is a narrowing-down condition candidate in which the ratio of the second number to the first number is within a predetermined range according to the first number and the operating load. Is a narrowing-down condition for notifying the driver.

  The audio output I / F 10 and the display 11 (informing means of the present invention) notify the driver of the narrow-down conditions determined by the narrow-down condition determining means 9 by voice output and screen display, respectively. The voice output I / F 10 outputs a voice guide by synthesizing voices, phrases, sentences, and the like stored in advance according to the extracted first number of search targets and filtering conditions. The display 11 displays on a screen the first number of extracted search targets, the narrowing conditions, the second number of narrowing conditions, or displays a list of the extracted search targets.

  Next, the operation (information search process) of the information search apparatus 1 of this embodiment will be described with reference to the flowcharts shown in FIGS. FIG. 2 is a flowchart showing the overall operation of the information search process (main routine process of the information search apparatus 1), and FIG. 3 is a flowchart showing the narrowing search process (subroutine process). In the following, as illustrated in FIG. 4A, a case where the driver searches for information by voice input “is there a delicious place around here?” Will be described as an example.

  With reference to FIG. 2, first, the first search condition is inputted by voice to the voice input I / F 3 from the driver of the vehicle 12 (STEP001). The voice input I / F 3 parses the driver's utterance input by voice by voice recognition processing, and acquires search conditions based on the keywords in the syntax. For example, as illustrated in FIG. 4A, the driver inputs a voice saying “Is there something delicious around here?” Then, the voice recognition process obtains the search condition “on the driving route and near the current position” from the keyword “this neighborhood”, and the search condition “restaurant” from the keyword “delicious tokoko”. Is done. Thus, the driver can easily input the search condition even during driving.

  Next, in STEP 002, the search target extraction unit 4 extracts a search target corresponding to the search condition from the search targets stored in the database 2. The search target extraction unit 4 determines a predetermined area (for example, the vehicle 12) on the map data from the current position of the vehicle 12, the travel route of the vehicle 12, and the search condition “on the travel route and close to the current position”. An area within a predetermined distance from the current position) is obtained, and a restaurant in the area is extracted. The current position and travel route of the vehicle 12 are acquired by a general method using GPS or the like. Then, the number of search targets (first number NA) extracted by the search target extraction unit 4 is calculated by the first number calculation unit 5.

  Next, in STEP003, it is determined whether or not the first number NA is not zero. If the determination result in STEP 003 is NO (the first number NA is 0), a predetermined voice guide that prompts re-input of search conditions is output from the voice output I / F 10. For example, as illustrated in FIG. 4B, the voice guide outputs “There is no restaurant around here. What do you want to do?” Then, returning to STEP 001, the driver's next utterance is inputted by voice.

  If the determination result in STEP 003 is YES (the first number NA is not 0), the narrowing-down condition candidate setting means 6 sets narrowing-down condition candidates that can be input from the voice input I / F 3. The narrowing-down condition candidates are predetermined in association with the search conditions. As a narrowing-down condition candidate, for example, when narrowing down the search target extracted by the search conditions “on the travel route and close to the current travel position” and “restaurant”, “genre” “dedication” “distance” “Store” is set.

  At this time, when the narrowing-down condition is “genre”, the contents that can be specified from the voice input I / F 3 in the information search apparatus 1 are, for example, three types of “Japanese food”, “Soba”, and “Family restaurant (family restaurant)”. is there. When the narrowing-down condition is “sticky”, there are two types of contents that can be designated from the voice input I / F 3 in the information search apparatus 1, for example, “individual management” and “franchise”. Further, when the narrowing-down condition is “distance” (distance from the current position of the vehicle 12), the contents that can be specified from the voice input I / F 3 in the information search apparatus 1 are “1 km”, “2 km”, and “5 km”. ”. Further, when the narrowing-down condition is “store” (store scale), the information search apparatus 1 can specify the contents that can be specified from the voice input I / F 3 in three types, for example, “large”, “medium”, and “small”. is there. In this way, the details of the narrowing conditions that can be specified are predetermined for each narrowing condition candidate. Of the candidates for narrowing down conditions, when special conditions such as “discretion” are used as narrowing conditions for general conditions such as “genre”, “distance”, and “store” As a result, the number of detection targets is greatly narrowed down.

  Next, in STEP 004, it is determined whether or not the first number NA is greater than a predetermined maximum number M. The maximum number M is the number of cases determined in advance as the maximum number of cases where information can be presented. If the determination result in STEP 004 is NO (the first number NA is equal to or less than the maximum number M), the process proceeds to STEP 006 as it is without performing the below-described narrowing search process, and information is presented.

  In STEP 006, in the case of presenting information when the narrow search process is not executed, a predetermined voice guide for notifying the search condition, the first number NA, and the like is output by the voice output I / F 10 and the display 11 As a result, the search condition, the first number NA, the candidate for the narrowing condition, and the like are displayed on the screen. For example, as illustrated in FIG. 4C, the voice guide outputs “NA restaurants in this area. Are you sure about detailed guidance?”. Further, for example, as illustrated in FIG. 5A, the number of restaurants and candidates for narrowing down conditions are displayed on the screen. Then, when the driver inputs a voice requesting detailed guidance, the list of extracted search objects is displayed on the display 11. Thereby, when there are few search objects extracted by the first search condition, the searched information is promptly presented.

  Next, if the determination result in STEP 004 is YES (the first number NA is greater than the maximum number M), the process proceeds to STEP 005, where a narrow search process for narrowing down the search target with more detailed search conditions (squeezing conditions) is performed. Executed.

  The narrowing search process is executed as shown in FIG. First, in STEP 101, the driving load detection means 8 detects the driving load of the driver based on the result of detecting the vehicle speed, steering operation, braking operation, palm sweating, and the like. The detected driving load is digitized. Next, in STEP 102, it is determined whether or not the operating load is higher than a predetermined threshold value. At this time, the utterance guidance mode flag is set to 0 in the initial state. The utterance guidance mode flag is a flag indicating whether or not it is a utterance guidance mode when the driving load is high, and is set to 1 when the driving load is high, and is set to 0 otherwise. If NO in STEP 102, the process proceeds to STEP 104 as it is. If YES in STEP 102, the driving load is high, the process proceeds to STEP 103, and the speech guidance mode flag is set to 1.

  Next, in STEP 104, the second number calculation means 7 calculates the second number NB. For example, when the filtering condition is “genre”, the number of cases nb1 when “Japanese food” is designated, the number of cases nb2 when “soba” is designated, and “family restaurant” are designated as the second number NB. Three values of the number of cases nb3 are calculated, and NB = {nb1, nb2, nb3}. Further, for example, when the narrowing-down condition is “sticky”, the second case number NB includes two values: the number nb4 when “individual management” is designated, and the number nb5 when “franchise” is designated. Is calculated, and NB = {nb4, nb5}. In this way, the number of search targets corresponding to the content of the extracted search targets (first number NA) for each content that can be specified with the narrowing condition is the second number NB of the narrowing condition. Calculated as an element.

  Next, in STEP 105, the narrow-down condition is determined by the narrow-down condition determining means 9, and the driver is notified by the audio output I / F 10 and the display 11. As a result, an utterance in which the driver specifies the narrow-down condition is guided. First, the narrowing-down condition determining unit 9 calculates the ratio X of the second number of cases NB to the first number of cases NA for each narrowing condition candidate. For example, when the narrow-down condition is “genre”, the ratio X = {nb1 / NA, nb2 / NA, nb3 / NA}. Further, when the narrowing-down condition is “adherence”, the ratio X = {nb4 / NA, nb5 / NA}. As described above, the ratio of the second number NB of the narrowing condition to the first number NA is calculated as an element of the narrowing condition ratio X. As a result, it is possible to know in advance how much the search target is narrowed down when a search is made by narrowing down with candidates for each narrowing condition.

  Next, the narrow-down condition determining unit 9 determines whether each element of the ratio X is included in the predetermined range R. And the narrowing-down condition determination means 9 determines the narrowing-down condition candidate in which all the elements of the ratio X are included in the range R as the narrowing-down conditions for notifying the driver. The range R is a range defined to be greater than the lower limit value Rmin and less than or equal to the upper limit value Rmax. Here, Rmin and Rmax are predetermined values satisfying 0 <Rmin <Rmax ≦ 1.

  In addition, as the range R, any one of three predetermined ranges R1, R2, and R3 is selectively used. The ranges R1, R2, and R3 are set such that the lower limit value Rmin and the upper limit value Rmax are lower as the level is higher. For example, in the first stage range R1, Rmin = 0.2 and Rmax = 0.5, and in the second stage range R2, Rmin = 0.1, so that Rmin and Rmax are lower than the first stage. Rmax = 0.2, and in the third stage range R3, Rmin = 0 and Rmax = 0.1 so that Rmin and Rmax are lower than in the second stage. In this example, the lower limit value Rmin and the upper limit value Rmax of the ranges R1, R2, and R3 are set so that the ranges R1, R2, and R3 do not overlap.

  The narrowing-down condition determining means 9 selects any one of the ranges R1, R2, and R3 as a range R depending on the magnitude relationship between the first number NA and two predetermined thresholds Th1 and Th2, and the value of the utterance guidance mode flag. Choose as. Here, Th1 and Th2 are predetermined values that satisfy M <Th1 <Th2 (M is the maximum number M described in STEP004). For example, Th1 is 20 and TH2 is 50.

  At this time, when the utterance guidance mode flag is 0 (the driving load is low), the range R1 is used as the range R when M <NA ≦ Th1, and the range R2 as the range R when Th1 <NA ≦ Th2. When Th2 <NA, the range R3 is used as the range R. Accordingly, the range R is determined such that the upper limit value Rmax is lower as the first number NA is larger, and the narrowing condition for notifying the driver is determined so that the second number NB is narrowed down.

  When the utterance guidance mode flag is 1 (the driving load is high), when M <NA ≦ Th1, a range R2 that is one step higher than when the driving load is low is used as the range R, and Th1 < When NA ≦ Th2, the range R3 is used as the range R, which is one step higher than when the operating load is low. When Th2 <NA, the range R3 is used as the range R. Thereby, when the driving load is high and it is necessary to concentrate by driving, a narrowing condition that can narrow down the search target is determined from the narrowing condition candidates.

  Next, a voice guide is output from the voice output I / F 10 in accordance with the determined narrowing condition, and the search condition, the first number NA, the narrowing condition, and the second number of narrowing conditions are displayed on the display 11. NB (number of items for each content that can be specified as a filtering condition) is displayed on the screen. In this way, since the second number NB of the narrowing condition is displayed on the screen together with the narrowing condition, the user can select the search target by the narrowing condition when specifying the narrowing condition. It can be grasped beforehand whether it is narrowed down.

  Here, examples of voice guidance and screen display will be described in detail with reference to FIGS. 4D, 4E, 5B, and 5C. First, the case where the voice guide is as shown in FIG. 4D and the screen display is as shown in FIG. 5B will be described as an example. In this case, the first number NA is 50, and the operation load is not high. As the range R, the range R2 is used. Then, when the filtering condition is “genre”, the elements {nb1 / NA, nb2 / NA, nb3 / NA} of the ratio X of the second number NB to the first number NA are all included in the range R2. Therefore, the “genre” among the candidates for the narrowing-down conditions is determined as the narrowing-down conditions to notify the driver.

  At this time, as exemplified in FIG. 4D, the voice guide outputs “NA restaurants in this area. Are there hopes for the genre?”. In addition, as illustrated in FIG. 5B, the screen shows the number of restaurants, narrowing conditions ("genre"), contents that can be designated for narrowing conditions ("Japanese food", "soba", "family restaurant"), The second number NB of the narrowing-down conditions (the number nb1 of “Japanese food”, the number nb2 of “soba”, the number nb3 of “family restaurant”) is displayed. And since the number of cases to be detected is narrowed down to 1/5 or less (10 cases or less) by the driver specifying the content of the “genre” notified in this way, the number of cases to be detected is up to the maximum number M. It is narrowed down efficiently.

  Next, the case where the voice guide is as shown in FIG. 4E and the screen display is as shown in FIG. 5C will be described as an example. In this case, the first number NA is 100 and the operation load is high, and the number of detection targets needs to be further narrowed down. Therefore, as the range R, the range R3 is used. In addition, when the narrowing-down condition is “sticky”, the elements {nb4 / NA, nb5 / NA} of the ratio X of the second number NB to the first number NA are all included in the range R3. “Feeling” among the candidate conditions is determined as a narrowing-down condition to notify the driver.

  At this time, as illustrated in FIG. 4E, the voice guide outputs “There are NA restaurants in this area. In addition, as illustrated in FIG. 5C, the number of restaurants, narrowing conditions ("sticking"), contents that can be specified for narrowing conditions ("individual management", "franchise"), narrowing are displayed on the screen. The second number of conditions NB (number of individual management nb4, number of franchise nb5) is displayed. And since the number of cases to be detected is narrowed down to 1/10 or less (10 cases or less) by the driver specifying the content of “notice” notified in this way, the maximum number of cases to be detected is M. It is narrowed down efficiently.

  Next, proceeding to STEP 106, it is determined whether or not to execute a narrow search. When the driver designates the narrowing condition, it is determined that the narrowing search is executed. When the driver does not designate the narrowing condition, it is determined that the narrowing search is not executed. If the determination result in STEP 106 is NO (no refinement search is performed), the refinement search process is terminated as it is, and the process returns to STEP 005 in FIG. 2 and proceeds to STEP 006, where the aforementioned refinement search process is not performed. Information is presented in the same way as information is presented.

  If the determination result in STEP 106 is YES (executes the refinement search), the process proceeds to STEP 107, where the refinement search is executed with the content of the refinement condition specified by the driver. For example, when “genre” is notified as a narrowing condition and “soba” in the content of “genre” is designated by voice input from the voice input I / F 3 by the driver, the search target extraction unit 4 sends the search target to the restaurant. Among the corresponding search targets, a search target corresponding to “soba” is extracted. Next, in STEP 108, it is determined whether or not the number of search targets extracted by the narrow search is equal to or less than the maximum number M. At this time, the optimum narrowing condition is notified to the driver in STEP 105, and the content of the optimum narrowing condition is designated by the driver. Therefore, the number of detection targets is usually narrowed down to the maximum number M or less. It is thought that there is. If the determination result in STEP 108 is YES, the narrow-down search process is terminated, the process returns to STEP 005 in FIG. 2, the process proceeds to STEP 006, and information is presented. In STEP 006, in the case of presenting information when the narrow search process is executed, the narrowed search targets are displayed as a list on the display 11. Thereby, the search object is efficiently narrowed down and the searched information is promptly presented.

  However, for example, when the first number NA is extremely large, the number of search targets cannot be sufficiently narrowed down, and the determination result of STEP 108 may be NO. In this case, the process returns to STEP 101 and the narrow-down search process is repeated. Thereby, even when there are an extremely large number of extracted detection targets, the search targets are efficiently narrowed down.

  Even if there are many extracted search targets by the above information search processing, the search targets are efficiently narrowed down. As a result, the information search apparatus 1 can appropriately present information while minimizing the number of searches and the time required for the search.

  Furthermore, in the present embodiment, when the driver selects a search target that he / she wants to know in detail from among the search targets displayed in a list, detailed information on the selected search target or the selected search target The travel route to may be guided by the audio output I / F 10 and the display 11.

  In the present embodiment, the driving load detection means 8 is provided, but the driving load detection means 8 is not provided, and the narrowing condition determination means 9 is based on the first number and the second number. You may make it determine the narrowing-down conditions notified to a driver | operator from the candidate of conditions. At this time, the narrowing-down condition determining means 9 operates the narrowing-down condition candidates in which the ratio of the second number of cases to the first number is within a predetermined range corresponding to the first number among the narrowing-down condition candidates. It is a narrowing condition to notify the person.

  In the present embodiment, the database 2 is a storage medium such as a CD-ROM, DVD, or HDD. However, an IC card, a magnetic card, or the like may be further used. Further, as the database 2, an information center that stores data necessary for navigation and is provided via a communication line according to a driver's request, an electronic notebook that stores data such as map data and destination data in advance, etc. An information source may be further used to exchange data via a wireless device.

  In this embodiment, the voice input I / F 3 is used as an input unit. However, an input I / F provided with a touch panel, a keyboard, buttons, a dial, and the like may be used. Further, the touch panel may be a display in which a touch switch is incorporated in a display which is a notification unit.

  In this embodiment, the voice output I / F 10 and the display 11 are used as notification means. However, a printer or the like that prints out the searched information may be further provided.

  In the present embodiment, the search target is a destination such as a restaurant. However, for example, information such as music may be the search target.

  In the present embodiment, the narrow-down condition determining unit 9 selects the range R from the three stages of ranges R1, R2, and R3, and when selecting the range R, the two types of threshold values Th1 and Th2 are set. Although used, the narrowing-down condition determining means 9 may be any one that selects the range R from a plurality of ranges using threshold values.

  In the present embodiment, the user who searches for information is the driver of the vehicle 12, but may be an occupant other than the driver.

  Moreover, in this embodiment, although the information search device 1 shall be mounted in the vehicle 12, it is good also as what is mounted in mobile bodies other than a vehicle.

The functional block diagram of the information search device which is this embodiment. The flowchart which shows the information search process of the information search device of FIG. The flowchart which shows the narrowing-down search process of the information search process of the information search device of FIG. FIG. 3 is a voice guide example in the information search apparatus of FIG. 1. FIG. The example of a screen in the information search device of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Information retrieval apparatus, 2 ... Database (memory | storage means), 3 ... Voice input I / F (input means), 4 ... Search object extraction means, 5 ... 1st number calculation means, 6 ... Narrowing condition candidate setting means, 7 ... 2nd number calculation means, 8 ... Driving load detection means, 9 ... Refinement condition determination means, 10, 11 ... Audio output I / F, Display (notification means), 12 ... Vehicle (moving body).

Claims (4)

Storage means for storing search objects;
Input means for inputting a search condition for searching for a search object stored in the storage means from a user;
Search object extraction means for extracting a search object corresponding to the search condition input to the input means;
First number calculation means for calculating a first number that is the number of search objects extracted by the search object extraction means;
Narrowing condition candidate setting means for setting candidates for narrowing conditions that can be input to the input means for narrowing down and searching the search objects extracted by the search object extracting means;
The number of search targets corresponding to the narrowing condition candidates among the search targets extracted by the search target extracting means for each of the narrowing condition candidates set by the narrowing condition candidate setting means. A second number calculating means for calculating the number of cases;
A means for determining the narrowing-down condition for notifying the candidate of the previous SL refinement condition to the user, of candidates for the narrowing-down condition, the ratio of the second count to the first count is, first count A narrowing condition determining means for determining a narrowing condition candidate existing in a predetermined range set according to the narrowing condition to be notified to the user ;
Informing means for notifying the user of the narrowing-down conditions determined by the narrowing-down condition determining means ,
The information search apparatus according to claim 1, wherein the predetermined range is a range set in accordance with the first number so that the upper limit of the range is lower as the first number is larger . An information retrieval device mounted on a mobile object,
  A driving load detecting means for detecting a driving load of the user who is a driver among the occupants of the mobile body;
  In addition to the upper limit of the range being lower as the first number is larger, the upper limit of the range is lower as the driving load detected by the driving load detecting means is higher. Thus, the information search device according to claim 1, wherein the range is set according to the first number and the operation load. The notifying means includes means for notifying the user of the second number of the narrowing conditions calculated by the second number calculating means together with the narrowing conditions determined by the narrowing condition determining means. The information retrieval apparatus according to claim 1 or 2 , characterized in that The entering force means, the information retrieval apparatus according to claim 2, wherein the search condition by the user of the speech is a passenger of said moving body is input.

Images