JP2006031149A - Information retrieving device and terminal equipment for information retrieval - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information retrieving device suitable for being mounted on a vehicle for generating the optimal category according to a retrieval result, and for proposing the target category of a user based on traveling information. <P>SOLUTION: This information retrieving device is provided with an input part 10 for inputting a retrieval request, a retrieving part 31 for retrieving information including the retrieval request according to the inputted retrieval request, a category deciding part 32 for deciding a category based on the retrieved information, an attribution level calculating part 33 for calculating the level of the attribution of the retrieved information to each category, word extracting parts 34 and 35 for deciding whether or not the extracted word is pertinent to a preliminarily registered word from the retrieved information, a category selecting part 36 for applying weighting to the extracted word, and for calculating scores for every category, and for selecting the category whose store is high and an output part 20 for outputting the selected category. Thus, it is possible to accurately propose a category desired by a user, and it is possible for the user to narrow down the results. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information search apparatus that can automatically narrow down a search result by automatically determining a classification system according to a search result, organizing and presenting the search result according to the classification system. .

  Conventionally, as this type of information retrieval apparatus, a document retrieval unit that retrieves a document database in accordance with a retrieval request and a plurality of documents retrieved by the document retrieval unit are designated in advance based on the similarity between the documents. There is known one having classification means for classifying documents into a number of document groups (categories), and an attribution degree calculation unit for calculating the degree to which each document obtained by the search belongs to each of the document groups. (For example, refer to Patent Document 1).

According to this, a plurality of categories representing the retrieved document set are automatically extracted by clustering, and for each retrieved document, the degree of belonging to each of the plurality of categories can be calculated, These attributions can be displayed on the screen, and for the category specified by the user, multiple searched documents can be sorted according to the attribution to the specified category. The search results can be overviewed by a category system suitable for the search, and the search results can be summarized by the category of interest.
JP 2003-345810 A

  However, in the conventional information search device described in Patent Document 1, when employed in an in-vehicle environment, the user needs to pay attention to the search results displayed on the display during driving and select them. There was a problem that this was practically difficult.

  The present invention has been made in view of such a conventional problem, and it is possible to generate an optimum category according to a search result and propose a category intended by a user based on traveling information. An object of the present invention is to provide an information search apparatus suitable for the above.

  An information search apparatus according to the present invention includes an input unit for inputting a search request, a search unit for searching for information including the search request in accordance with the input search request, and a category based on information searched by the search unit. Category determination means for determining the degree of attribution of the searched information belonging to each category, and each word extracted from the searched information corresponds to a pre-registered word A word extracting means for determining whether or not to extract, a category selecting means for assigning a weight to the word extracted by the word extracting means, calculating a score for each category, and selecting the category having a high score; And an output means for outputting the category selected by the category selecting means.

  With this configuration, according to the searched information, each category is determined, a score is calculated, and a category having a high score can be automatically proposed as a category desired by the user. The desired category can be easily narrowed down without sequentially gazing at information.

Further, the information search apparatus of the present invention is characterized in that the word extraction means determines whether each word extracted from the searched information corresponds to a registered facility, and the searched information And a place name extracting means for judging whether or not each word extracted from the word corresponds to a registered place name.

  With this configuration, the facility and place name are extracted by the word extraction means, the extracted facility and place name are weighted, and the score of each category is calculated based on the extracted place. It has the effect that more place names are included.

  The information search apparatus of the present invention further includes positioning means for positioning the current position, destination acquisition means for acquiring a destination, and path acquisition means for acquiring a route to the destination, the category selection means. However, for the words extracted by the word extraction means, the distance between each of the words and the current position, the destination, and the position of the route acquired by the positioning means, the destination acquisition means, and the route acquisition means It has the structure which performs weighting according to.

  With this configuration, the weights for the words extracted by the word extraction means are close to the current position, the destination, and the path position according to the distance between the word position and the current position, the destination, and the path position. The effect is that the weighting increases.

  Further, the information search device of the present invention includes positioning means for positioning the current position and the moving direction thereof, and the category selecting means for the word extracted by the word extracting means, the position of the word, It has the structure which weights according to the present position and the moving direction which were measured by the positioning means.

  With this configuration, even in the vicinity of the current position, a word that is close to the moving direction can be heavily weighted, and a category including more words in the moving direction can be proposed.

  Further, in the information search apparatus of the present invention, the category selecting unit calculates a score for each category by multiplying the weight extracted to the word extracted by the word extracting unit and the number of times of extraction, It has the structure which selects the said category with a high score.

  With this configuration, the score for each category is calculated by multiplying the weight assigned to the word and the number of extractions, and the category that appears more frequently and includes the more weighted word is proposed first. Has the effect of

  The information search apparatus according to the present invention further includes a dialog control unit that can be controlled by a dialog, and has a configuration in which an input from the input unit and an output to the output unit are performed in a dialog format by the dialog control unit.

  With this configuration, it is possible to perform interactive processing from the search request until the final result is output, and the operation is facilitated.

  The information search apparatus according to the present invention has a configuration in which the dialog control unit includes a voice recognition unit and a voice synthesis unit, respectively, so that the input from the input unit and the output from the output unit can be performed by voice.

  With this configuration, input and output can be performed by voice, and the operation can be easily performed even when the user cannot keep an eye on the vehicle-mounted device.

The information search device of the present invention includes a movable terminal device and a center device capable of communicating with the terminal device, and the terminal device inputs the search request, and each category. A category selecting unit that calculates the score and selects the category having a high score; and the output unit that outputs the category selected by the category selecting unit. Search means for searching for information including the search request according to the input search request, category determining means for determining a category based on the information searched by the search means, and the searched information for each category A degree-of-attribution calculating means for calculating the degree of belonging;

  With this configuration, a search request can be made from a movable terminal device, a search can be executed by the center device, and a desired category can be selected and output by the terminal device based on the result. It has the effect of not requiring a database.

  Furthermore, the information search terminal device of the present invention calculates a score for each category determined by the center device in accordance with the input means for inputting the search request and the search request input by the input means. , Category selecting means for selecting the category having a high score, and output means for outputting the category selected by the category selecting means.

  With this configuration, there is an effect that a desired category can be easily narrowed down without sequentially looking at information searched for by the user.

  The information search terminal device of the present invention further includes positioning means for positioning the current position, destination acquisition means for acquiring the destination, and path acquisition means for acquiring the route to the destination. Have

  With this configuration, the weights for the words extracted by the word extraction means are close to the current position, the destination, and the path position according to the distance between the word position and the current position, the destination, and the path position. The effect is that the weighting increases.

  Moreover, the terminal device for information retrieval of this invention has the structure by which the said terminal device is mounted in a vehicle.

  With this configuration, the effect is great when the display unit is mounted on a vehicle and the display unit cannot be watched.

  An information search apparatus according to the present invention includes an input unit for inputting a search request, a search unit for searching for information including the search request in accordance with the input search request, and a category based on information searched by the search unit. Category determination means for determining the degree of attribution of the searched information belonging to each category, and each word extracted from the searched information corresponds to a pre-registered word A word extracting means for determining whether or not to extract, a category selecting means for assigning a weight to the word extracted by the word extracting means, calculating a score for each category, and selecting the category having a high score; , Output means for outputting the category selected by the category selection means, determine the category according to the searched information, The core can be calculated, and a category with a high score can be automatically proposed as a category desired by the user, and the desired category can be easily narrowed down without sequentially looking at the information searched by the user. Has an effect.

  Hereinafter, an information search apparatus according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram of an information search apparatus according to Embodiment 1 of the present invention.

  As shown in FIG. 1, an information search apparatus according to an embodiment of the present invention is movable, for example, a vehicle-mounted terminal device 100, and a center device 200 that can communicate with the terminal device 100 via a network 300. It consists of and.

  The terminal device 100 includes an input unit 10, an output unit 20, a control unit 30a, a positioning unit 39, and a database 40.

  The input unit 10 includes any one or more of a remote controller, a microphone, a touch panel, a keyboard, and the like, and inputs a keyword indicating contents to be searched for, inputs a destination, and other commands and information. Can be input.

  The output unit 20 includes a display, a speaker, and the like, and appropriately displays and outputs information from the input unit 10, results and information calculated by the control unit 30a, and the like.

  The control unit 30a searches for desired information based on information and instructions from the input unit 10, and automatically determines and presents the most desired information from the search result, which will be described later. It has a configuration to do.

  The positioning unit 39 measures the current position of the vehicle on which the terminal device 100 is mounted, the moving direction thereof, and the like using a GPS (Global Positioning System) receiver, a self-contained navigation positioning device, or the like.

  The database 40 includes a map database 41 having map information, a facility name database 42 having facility names appearing on the map, and a place name database 43 having place names appearing on the map.

  Here, the control unit 30a described above transmits the information and command (search request) input from the input unit 10 to the center device 200 and receives information from the center device 200, and the center. Of the words that appear in the information received from the apparatus 200, if there is a word that corresponds to the facility name, the facility name extraction unit 34 that extracts the facility name from the facility name database 42, and among the words that appear in each information If there is a place name corresponding to the place name, the place name extraction unit 35 for extracting the place name from the place name database 43, and the words (facility name, place name) extracted by the facility name extraction unit 34 and the place name extraction unit 35 in advance A category selection unit 36 that assigns weights according to defined rules, calculates the scores of each category based on the weights, and arranges them in descending order of scores, and an input When a destination is input from 10, a destination acquisition unit 37 that acquires the destination, a route acquisition unit 38 that acquires a route to the destination, and a dialog that can perform speech recognition, speech synthesis, or the like And a control unit 50.

  On the other hand, the center apparatus 200 includes a control unit 30b and an information database 44.

  The control unit 30b receives the information and command (search request) transmitted from the terminal device 100, and transmits the search result, category, attribution level, and the like, and the information received by the communication unit 70. The search unit 31 that searches the information database 44 for information in accordance with the command (search request) and the result of the search by the search unit 31 are automatically determined according to the result. A category determining unit 32 for calculating the degree of attribution of each piece of information to each category determined by the category determining unit 32.

  As a method for automatically determining a category set in the category determination unit 32, an existing clustering method as described in Patent Document 1 may be used.

  By the way, the facility name extraction unit 34 and the place name extraction unit 35 of the control unit 30a constitute a word extraction unit for extracting a word (facility name, place name), respectively, and the words extracted by these extraction units 34 and 35. When weighting (facility name, place name), weighting is performed as follows using the information of the destination acquisition unit 37, route acquisition unit 38, and positioning unit 39, respectively.

  First, the destination acquisition unit 37 acquires information such as an address, a place name, a facility name, and latitude / longitude for the destination input from the input unit 10, and the route acquisition unit 38 acquires the destination from the current position. The route to the ground is searched and the route information is acquired.

  That is, when the destination is set, weighting is performed based on the destination, the route, and the distance from the current position acquired by the destination acquisition unit 37, the route acquisition unit 38, and the positioning unit 39. Is not set, weighting is given from the current position measured by the positioning unit 39 and the moving direction based on the distance in that direction.

  Further, as shown in FIG. 2, the category selection unit 36 determines how far the location of the facility name and place name is from the current position, the destination, and the route when the facility name and place name are extracted. Sort the categories in the order of the calculated score, the distance calculating unit 361 that calculates on the map, the word weighting unit 362 that assigns weights according to the calculated distance, the category score calculating unit 363 that calculates the score of the category, Category sorting unit 364, information score calculating unit 365 that calculates the score in the same way for information, information sorting unit 366 that sorts information in the order of the calculated score, and category information by integrating these information. A category information acquisition unit 367 that acquires the information and outputs it to the output unit 20. Further, the output unit 20 is configured to output to the user the category that is calculated first and has the highest score.

  Next, the operation of the information search apparatus in the present embodiment will be described.

  FIG. 3 is a flowchart showing the operation of the information search apparatus in the present embodiment. 5 to 10 (a) to 10 (c) are conceptual diagrams showing examples of specific operation states during the operation.

  First, as shown in FIG. 3, a keyword indicating an arbitrary search request is input using the input unit 10 (S301). Then, the input search request is supplied to the control unit 30a, and is transmitted to the center apparatus 200 via the communication unit 60 under the control of the control unit 30a.

  In the center device 200, the search request transmitted from the terminal device 100 is received by the communication unit 70 and input to the search unit 31. The search unit 31 searches the information database 44 for related information including the search request (keyword) based on the input search request (keyword) <S302>.

  The information database 44 stores a large amount of information related to the input search request (keyword). For example, as shown in FIG. 4, when the keyword “XXX” is input, A plurality of documents A, B, C, D,... Are searched as information including the keyword “OO”.

  When a plurality of documents A, B, C, D,... Are searched as information including the keyword “XX”, which of these documents A, B, C, D,. The category determining unit 32 automatically determines the category set to determine whether the category is satisfied (S303).

When the category determining unit 32 automatically determines a category set, for example, an existing hierarchical bottom-up clustering method may be used. In the hierarchical bottom-up clustering method, a cluster in which each piece of information consists only of itself is created as an initial state. Here, each piece of information (cluster) is represented by a vector whose element is an appearance word indicated by a black square in FIG. A word that is each element of the vector is weighted based on the appearance frequency. At that time, the word having the largest weight is set as the category name of the cluster set. Then, the distance between clusters is calculated for all cluster pairs. As the distance, the cosine between vectors is often calculated. Among all cluster pairs, the shortest distance cluster values are merged into one cluster. The merged cluster is also a vector having words as elements. Next, the distance information is updated by calculating the distance between the merged cluster and each remaining cluster.

  When the category set is determined in this way, the degree-of-assignment calculation unit 33 then calculates to what degree each information belongs to each category (S304). When calculating the degree of belonging, the distance (cosine) between each category and each information is calculated using a word vector, and the value is used as the degree of belonging to each category. The black circles in FIG. 4 indicate category vectors.

  FIG. 5 shows a state in which category sets are determined in this way, and how the information (documents A, B, C, D,...) Of each category set belongs to each category is calculated. Yes. For example, category 1 is a company name, and is a category “XX company”. This category 1 indicates that the document A as the retrieved information exists as one of the category set, and 0.9 belongs to this category 1. Category 2 is a product name, which is a category “XX product”. This category 2 indicates that the document D as the retrieved information exists as one of the category sets, and 0.8 belongs to this category 2. Category 3 is an individual name and is a category of “Mr. XXX”. This category 3 indicates that the document C as the retrieved information exists as one of the category sets, and 0.9 belongs to this category 3. The documents drawn behind the documents A, D, and C are one of the category sets belonging to the categories 1, 2, and 3, respectively, and belong to the respective categories 1, 2, and 3 at the respective degrees. It shows that you are doing.

  When the degree of attribution is calculated in this way, this information is transmitted to the terminal device 100 via the communication unit 70 together with the retrieved information (documents A, B, C, D,...). In the terminal device 100, the information is received by the communication unit 60, and the facility name extraction unit 34 is used for each piece of information (documents A, B, C, D,...) A word corresponding to the name is extracted, and the number of extractions is calculated (S305). Here, when the facility name is extracted, the facility name database 42 is used, and the facility name registered in the facility name database 42 and the appearance in the searched information (documents A, B, C, D,...). When there is a match by comparison with a word, the word is extracted from the facility name database 42.

  Similarly, using the place name extraction unit 35, for each piece of information (documents A, B, C, D,...), The word corresponding to the place name is extracted from each place (category A, B, C, D,...) The number of extractions is calculated (S306).

  FIG. 6 shows a state in which facility names and place names are extracted for each category and the number of extractions is calculated.

When the words (facility name and place name) are extracted in this way, the category selection unit 36 next checks whether or not the destination is input and set in advance (S307). When the destination is set (Yes in S307), the distance and the longitude of the extracted word (facility name and place name) in the distance calculation unit 361 constituting the category selection unit 36 are the destination, current position, First of all, the distance from the route to the destination is calculated. Then, based on the calculated distance, the word weighting unit 362 gives weights to the extracted words (facility name and place name), respectively (S308). When weighting is given, as shown in FIG. 7, the distance (d2, d1, d3) between the point indicated by the extracted word (facility name and place name) and the route to the destination, current location, and destination ) Is closer, the word weighting value is increased. Further, the weighting value is made smaller for the word located behind the current position with respect to the destination. However, the weighting value is in the range of 0 to 1.

  If the destination is not set <No in S307>, the extracted word is weighted based on the moving direction and the distance (d1) from the current position (S309). At that time, the weight of the word is increased as the point indicated by the extracted word (facility name and place name) is closer to the moving direction.

  After weighting words (facility names and place names), based on the weighted words (facility names and place names), the category score calculation unit 363 calculates the total score for each category as shown in FIG. calculate. That is, as a calculation method, the number of appearances of each extracted word (facility name and place name) is multiplied by the weight of each word (facility name and place name), and the sum of them is calculated. Then, based on the calculated total score, the category sorting unit 364 arranges categories in descending order of the total score (S310).

  On the other hand, the retrieved information (documents A, B, C, D,...) Is also shown in FIG. 9 in the information score calculation unit 365 based on the degree of association with the search request and the degree of belonging to each category. As described above, a score is calculated for each piece of information (documents A, B, C, D,...). Then, the information (documents A, B, C, D,...) Is arranged in descending order of score (S311). Then, in order to inform the user of the category with the highest score, the category information acquisition unit 367 acquires the information and outputs it to the output unit 20 (S312).

  The output unit 20 includes a display device, a touch panel, and the like, and displays category names in descending order of score as shown in FIG. Accordingly, the Yes / No button is operated for the upper category from the displayed categories, and the desired category is selected.

  In the case of No, as shown in FIG. 10B, the Yes / No button is operated in the same manner for the next category, and this is repeated.

  When a desired category is selected in this way, information (documents) in that category is displayed in descending order of score, as shown in FIG. Therefore, the desired information is selected and the details are displayed as before.

  As described above, according to the present embodiment, categories and information are displayed in descending order of scores, so that desired information can be easily displayed and output. That is, according to the present embodiment, the information search result is classified for each category according to the content of the search result, and only the category having the highest relationship with the destination, current position, route, etc. is selected from the categories. Therefore, the user does not have to bother to list the search results and narrow down each time, and when used as an in-vehicle information search device, the user can easily obtain desired information based on the driving situation Has the effect of being able to.

In this embodiment, word weighting is only calculated based on the destination, current position, route position, and distance, but if the passenger is a friend, a family member, or a child. Depending on the situation, the weight may be changed, the weight closer to the destination may be greater than the point closer to the current position, or vice versa. It is also possible to enlarge it. In addition, facility usage fees and user preferences may be added to the weighting conditions. In addition, by accumulating the destination setting history and the moving history, the user's frequently visited location is grasped, and the weighting is performed accordingly. May be changed.

  Further, in the present embodiment, in the description so far, input and output are described as being performed only by manual operation and display, but these can also be performed using sound. That is, in the present embodiment, as shown in FIG. 1, the control unit 30a constituting the terminal device 100 is provided with a dialog control unit 50 capable of voice recognition, voice synthesis, etc., and this dialog control unit 50 is used. For example, input and output can be performed not only by manual operation and display but also by voice.

  FIG. 11 is a flowchart showing main operations when input or output is performed using the dialogue control unit 50. Next, input and output operations when the dialog control unit 50 is used will be described using this flowchart.

  The dialogue control unit 50 automatically requests the user to input some request. The request is made by voice or display from the output unit 20. In order to perform input and output by voice, it is possible to use voice synthesis and voice recognition technology.

  When the dialog control unit 50 requests to input a predetermined request, the user inputs a desired request according to the request (S1113). Of course, this request can be input using a microphone, a keyboard, a touch panel, a remote controller, or the like, but this time, for example, is input by voice using a microphone. The request input by voice is recognized by the dialogue control unit 50, and it is first determined whether or not the request is a search request (S1114). If the request is a search request (No in S1114), the process proceeds to S302 shown in FIG. 3, and the operation as described above is executed. If the request is not a search request (Yes in S1114), it is determined whether the request is a category name previously searched as a search result (S1115). As a result of the determination, if it is not a category name (Yes in S1115), it is further determined whether or not the request is a voice “Yes” (S1117). That is, it is determined whether the voice request (keyword) corresponds to “Yes”, “No”, or the category name. In other words, the user selects one of the category names displayed on the screen or pronounced by voice, and checks whether or not it is pronounced.

  If the input request corresponds to the category name (No in S1115), the corresponding category is selected (S1116), and the information with the highest score in the category is output by voice (S1119). Then, a list of all information in the category is displayed on the screen in descending order of score (S1120). The screen image at that time is shown in FIG.

  If the input request does not correspond to the category name (Yes in S1115), it is determined whether the input request is “Yes” or “No” (S1117). If it is determined that a request indicating “No” has been input (No in S1117), the category having the next highest score after the currently presented category is presented as a candidate (S1118). The image of the screen at that time is FIG. If it is determined that a request indicating “Yes” has been input (Yes in S1117), the information with the highest score among the currently presented category information is output by voice (S1119), All information in the category is displayed in a list in descending order of score (S1120). The screen image at that time is as shown in FIG.

As described above, according to the present embodiment, since the dialogue control unit 50 is provided, information is presented according to the category classified according to the search result, and the content can be further interactively narrowed down. Even when the presented category is different from the category desired by the user, the category of the search result can be easily changed, and the desired information can be acquired.

  In the present embodiment, only one search request is input. However, two or more search requests may be input in the form of AND or OR. On the other hand, an AND or OR condition may be further added so that one or more new search requests can be input. Moreover, you may make it change the operation | movement of the dialog control part 50 dynamically with a user or a driving | running | working state.

  Moreover, according to this Embodiment, since it is comprised with the terminal device 100 which can be moved, and the center apparatus 200, the terminal device 100 which can be moved does not require the information database which stores much information, Necessary information can be arbitrarily obtained from the center apparatus 200 when necessary and output to the output unit 20.

  In the present embodiment, the center device 200 is provided with the communication unit 70, the search unit 31, the category determination unit 32, the attribution calculation unit 33, and the information database 44, but the center device 200 further includes the terminal device 100. A management unit for each terminal device 100, a category selection unit 36, a destination acquisition unit 37, a route acquisition unit 38, a facility name extraction unit 34, and a place name extraction unit 35 for each terminal device 100 are provided in the management unit. Only the final result may be received by the communication unit 60, and necessary information may be output to the output unit 20 via the dialogue control unit 50.

  In short, it is only necessary to appropriately set how far the functions are handled by the terminal device 100 and how much function the center device 200 is burdened with. That is, if the information that can be used is limited, the terminal device 100 can have all of these functions.

  The information search device of the present invention can determine a category according to the searched information, calculate a score, and automatically propose the category as a category desired by the user in descending order of the score. In this way, the information retrieved by the user is closely watched, and it is possible to easily and automatically narrow down the desired category without narrowing down the desired category. It is particularly useful as an information retrieval device that is difficult to do.

The block diagram of the information search device in one embodiment of this invention 1 is a block diagram of the main part of an information search apparatus according to an embodiment of the present invention. The flowchart which shows operation | movement of the information search device in one embodiment of this invention. The conceptual diagram explaining the operation | movement at the time of searching the information of the information search device in one embodiment of this invention, and determining a category The conceptual diagram which shows how the searched information of the information search device in one embodiment of this invention belongs to each category The conceptual diagram which showed how many times a facility and a place name appear in the information which belongs to each category of the information search device in one embodiment of this invention The conceptual diagram explaining the method of assigning | weighting to the facility of the information search device in one embodiment of this invention, and a place name The conceptual diagram explaining the operation | movement in the case of calculating the total score of each category of the information search device in one embodiment of this invention, and sorting The conceptual diagram which shows the state which calculated the score based on the relevance degree and the degree of attribution about each information of the category with the largest score of the information search device in one embodiment of the present invention and sorted (A) The figure which shows the result output state which guides the category 1 of the information search device in one embodiment of this invention (b) The figure which shows the result output state which guides the category 2 (c) The information result of the category 1 Diagram showing output status The flowchart which shows operation | movement of the dialog control part of the information search device in one embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Input part 20 Output part 30a, 30b Control part 31 Search part 32 Category determination part 33 Attribution degree calculation part 34 Facility name extraction part 35 Place name extraction part 36 Category selection part 37 Destination acquisition part 38 Path acquisition part 39 Positioning part 40 Database 41 Map Database 42 Facility Name Database 43 Place Name Database 44 Information Database 50 Dialogue Control Unit 60, 70 Communication Unit 100 Terminal Device 200 Center Device 300 Network 361 Distance Calculation Unit 362 Word Weighting Unit 363 Category Score Calculation Unit 364 Category Sorting Unit 365 Information Score Calculation unit 366 Information sorting unit 367 Category information acquisition unit

Claims (11)

Input means for inputting a search request; search means for searching for information including the search request according to the input search request; category determining means for determining a category based on information searched by the search means; A degree of attribution calculating means for calculating the degree to which the retrieved information belongs to each category, and a word to be extracted by determining whether each word extracted from the retrieved information corresponds to a pre-registered word An extraction unit, a weight selection unit for weighting the words extracted by the word extraction unit, calculating a score for each category, and selecting the category having a high score; And an output unit for outputting the category. The word extracting means is a facility extracting means for determining whether each word extracted from the searched information corresponds to a registered facility, and each word extracted from the searched information is registered. The information search apparatus according to claim 1, further comprising a place name extraction unit that determines whether the place corresponds to the place name. A word extracted by the word extracting means by the category selecting means, further comprising positioning means for measuring the current position, destination acquiring means for acquiring the destination, and route acquiring means for acquiring the route to the destination. On the other hand, weighting is performed according to the distance between the position of each word and the current position, the destination, and the position of the route acquired by the positioning unit, the destination acquisition unit, and the route acquisition unit. The information search device according to claim 1 or 2. Positioning means for positioning the current position and its moving direction, respectively, the category selecting means for the word extracted by the word extracting means, the position of the word, the current position and the moving direction measured by the positioning means 3. The information search apparatus according to claim 1, wherein weighting is performed according to the information. The category selection means calculates a score for each of the categories by multiplying the weights assigned to the words extracted by the word extraction means and the number of extractions thereof, and selects the category having a high score. 5. The information search apparatus according to claim 1, wherein the information search apparatus is characterized in that: 6. The apparatus according to claim 1, further comprising dialog control means that can be controlled by dialog, wherein the input from the input means and the output to the output means are each performed interactively by the dialog control means. The information search device described in 1. 7. The information search apparatus according to claim 6, wherein the dialogue control means includes a voice recognition means and a voice synthesis means, respectively, so that the input from the input means and the output from the output means can be made by voice. The terminal device has a movable terminal device and a center device communicable with the terminal device, and the terminal device calculates the score for each category, the input means for inputting the search request, and the score is high. A category selection means for selecting the category; and the output means for outputting the category selected by the category selection means, wherein the center device requests the search request according to the search request input by the input means. Search means for searching for information including information, category determination means for determining a category based on information searched by the search means, and degree-of-attachment calculation means for calculating the degree to which the searched information belongs to each category The information search apparatus according to claim 1, wherein the information search apparatus includes: The input unit for inputting the search request, and a category selection for calculating a score for each category determined by the center device according to the search request input by the input unit, and selecting the category having a high score And a terminal device for information retrieval comprising: the output means for outputting the category selected by the category selection means. 10. The information search terminal device according to claim 9, further comprising positioning means for measuring a current position, destination acquisition means for acquiring a destination, and route acquisition means for acquiring a route to the destination. The terminal device for information retrieval according to claim 9, wherein the terminal device is mounted on a vehicle.

Images