JP2023055839A - Destination conjecture device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To set a destination of path finding while paying no attention to operation for inputting the destination.

SOLUTION: A destination conjecture device includes: a retrieval information acquisition section for acquiring retrieval information including information on a retrieval term retrieved in a retrieval site; an attribution granting section for granting attribution information representing attribution of a retrieval term to the retrieval term; a retrieval term evaluation section for obtaining frequency of appearance of a retrieval term whose attribution is in common for each attribution; and a conjecture destination candidate extraction section for extracting a retrieval term related to a destination among retrieval terms included in attribution whose frequency of appearance satisfies a predetermined criterion as a conjecture destination candidate.

SELECTED DRAWING: Figure 1

COPYRIGHT: (C)2023,JPO&INPIT

Description

The present invention relates to a destination estimation technique for estimating a destination set by a user.

In route search services provided by in-vehicle navigation devices, websites, etc., it is necessary to set a destination when searching for a route. In general, the setting of the destination for route search is performed by receiving the input of the destination from the user. Destinations can be entered using a variety of methods, including name input, address input, phone number input, designation on a map, and selection from destination setting history. is also used to input the destination.

Further, as described in Patent Document 1, a device other than a route search device, such as a mobile terminal device, is used to input a destination, and the input destination is transferred to the route search device. is also proposed.

JP 2014-145751 A

As described above, various methods can be used to input a destination when searching for a route, and new methods have also been proposed. This is considered to indicate that the conventional task of inputting a destination is not necessarily easy and bothers the user, and that a method for reducing this burden is sought.

If the user avoids the work of inputting the destination to set the destination, the chances of using the route search service decrease, and the user cannot enjoy the convenience of the route search, which is undesirable. Therefore, if the user can set the destination of the route search without being conscious of the task of inputting the destination, the route search service will become more familiar, and it is expected that the opportunities for using the route search service will increase.

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to enable a user to set a destination for a route search without being conscious of the task of inputting the destination.

In order to solve the above-mentioned problems, the destination estimation device according to the first aspect of the invention includes a search information acquisition unit that acquires search information including information related to a search term searched on a search site; an attribute assigning unit that attaches attribute information indicating an attribute of a search term; a search term evaluation unit that obtains, for each attribute, the frequency of appearance of search terms having the same attribute; and an inferred destination candidate extraction unit that extracts, as an inferred destination candidate, a search term related to the destination from among the search terms.

In order to solve the above-mentioned problems, a destination estimation method according to a seventh aspect of the invention is a destination estimation method using an information processing device, in which search information including information related to a search term searched on a search site is obtained. a step of attaching attribute information indicating an attribute of the search term to the search term; a step of obtaining, for each attribute, the frequency of appearance of search terms having the same attribute; and an inferred destination candidate extraction step of extracting, as an inferred destination candidate, a search term related to a destination from among search terms included in attributes that satisfy a criterion.

In order to solve the above problems, the computer program according to the eighth aspect of the invention includes a search information acquisition unit for acquiring search information including information on a search term searched on a search site, An attribute assigning unit that attaches attribute information indicating an attribute, a search term evaluation unit that obtains, for each attribute, the frequency of appearance of search terms having the same attribute in common, and among search terms included in attributes whose frequency of appearance satisfies a predetermined criterion. , and an inferred destination candidate extraction unit that extracts a search term related to a destination as an inferred destination candidate.

In order to solve the above-mentioned problems, a recording medium according to the ninth aspect of the invention is a recording medium readable by an information processing apparatus, on which the computer program according to the eighth aspect is recorded.

1 is a block diagram showing a configuration example of a route search system to which a destination estimation device that is one embodiment of the present invention is applied; FIG. It is a figure which shows the equipment configuration example of a route search system. It is a figure which shows the equipment configuration example of a route search system. FIG. 10 is a flowchart for explaining an operation for extracting estimated destination candidates; FIG. It is a figure explaining grouping by classification of a search term. 4 is a flowchart for explaining an operation for determining an estimated destination;

An embodiment of the present invention will be described. A destination guessing device according to an embodiment of the present invention includes a search information acquisition unit that acquires search information including information on a search term searched on a search site; , a search word evaluation unit that obtains the frequency of appearance of search words having the same attribute for each attribute, and a search word included in the attributes whose appearance frequency satisfies a predetermined criterion. and an inferred destination candidate extraction unit that extracts a search term related to as an inferred destination candidate.

Since the user only needs to select a destination from among the extracted estimated destination candidates, the user is freed from the task of inputting the destination.

wherein the search information includes information on a date and time when the search word is searched, and an estimated destination determination unit that determines an estimated destination from the estimated destination candidates based on the information on the date and time. good too.

Alternatively, a position information acquisition unit that acquires position information indicating the user's position for each time, and an estimated destination determination unit that determines an estimated destination from the estimated destination candidates based on the position information. good too.

Further, the apparatus may include a preference information processing unit that processes user preference information related to user preferences, and an estimated destination determination unit that determines an estimated destination from the estimated destination candidates based on the user preference information. good.

Further, the search word evaluation unit obtains a temporal appearance tendency of search words for each classification based on the date and time information, and the estimated destination determination unit determines the appearance tendency of the estimated destination candidate in advance. It may be determined whether or not it matches a defined pattern.

In addition, in relation to the position information and the search word, the user characteristic evaluation unit classifies the behavior characteristics of the user into predetermined patterns, and the estimated destination determination unit determines the A guessed destination may be determined from the guessed destination candidates.

A method of estimating a destination according to an embodiment of the present invention comprises the steps of acquiring search information including information related to a search term searched on a search site, and attaching attribute information indicating an attribute of the search term to the search term. obtaining, for each attribute, the frequency of appearance of search words having a common attribute; and guessing destination candidates for search words related to a destination among the search words included in the attributes whose appearance frequencies satisfy a predetermined criterion. and an inferred destination candidate extraction step of extracting as

A computer program according to an embodiment of the present invention includes a search information acquisition unit that acquires search information including information about a search term searched on a search site, and attaches attribute information that indicates an attribute of the search term to the search term. an attribute assigning unit, a search term evaluation unit that obtains, for each attribute, the appearance frequency of search terms having a common attribute, and guesses a search term related to a destination among search terms included in attributes whose appearance frequency satisfies a predetermined criterion. The information processing device is caused to function as an inferred destination candidate extraction unit that extracts destination candidates.

A recording medium that is an embodiment of the present invention is a recording medium that records the computer program described above.

FIG. 1 is a block diagram showing a configuration example of a route search system 10 to which a destination estimation device according to one embodiment of the present invention is applied. As shown in the figure, the route search system 10 includes a destination estimation device 100 , a search device 210 , a position information acquisition device 220 , an estimated destination presentation device 230 and a route search device 240 .

The search device 210 is an information processing device that can use the Internet 310 and includes a web browser section 211 and a search list creation section 212 . The web browser unit 211 can be composed of general-purpose web browser software. In this embodiment, the user uses the Web browser unit 211 to access the general-purpose search site 300 and perform a keyword search. Since the search device 210 is an information processing device that a user uses on a daily basis, the keywords used for searching include not only keywords related to route search, but also miscellaneous keywords related to hobbies, life, work, and the like.

The search list creating unit 212 acquires the keywords used for the search by the web browser unit 211 accessing the search site 300 and saves them in a list format. However, the storage of keywords is not limited to the list format. The search list shall include information on the date and time of the search as well as the keywords. The search list creating unit 212 can be implemented as add-on software for the web browser unit 211, for example. Since the URL of the general-purpose search site 300 is known, the keywords entered at the search site 300 can be easily obtained.

When a linked web page is displayed as a search result, the search list may include the time when the web page was browsed.

The position information acquisition device 220 is a device that moves together with the user, and acquires position information indicating the user's position at each time using a GPS (Global Positioning System) function, access point information, etc., and attaches date and time information. Record as a location information list. However, the storage of location information is not limited to the list format. Specifically, it can be composed of a mobile phone, a smart phone, an in-vehicle navigation device, or the like.

The destination guessing device 100 is a device for guessing the destination desired by the user based on the search list created by the search device 210, and is equipped with a CPU, a memory, a communication device, etc., and information capable of executing a computer program. It can be constructed using a processing device. Since the destination guessing device 100 guesses the destination based on the search information on the search site 300, the user does not need to input the destination.

The destination estimation device 100 includes a search list acquisition unit 110 , a location information acquisition unit 120 , an estimated destination candidate extraction unit 130 , a user characteristic evaluation unit 140 and an estimated destination determination unit 150 .

The search list acquisition unit 110 acquires a search list from the search device 210 at a predetermined timing. A plurality of retrieval devices 210 may be acquired. However, the data to be acquired is not limited to the list format. The position information acquisition unit 120 acquires the position information list from the position information acquisition device 220 at a predetermined timing. A plurality of location information acquisition devices 220 may be acquired. However, the data to be acquired is not limited to the list format.

The estimated destination candidate extraction unit 130 is a block that extracts candidates for the estimated destination based on the search list acquired by the search list acquisition unit 110. It has an evaluation unit 133 , an estimated destination candidate extraction unit 134 , and a dictionary DB 135 .

The search term extraction unit 131 performs processing for extracting search terms from the keywords recorded in the search list. The date and time information of the extraction source keyword is also inherited to the extracted search term. That is, the keyword input by the user is not necessarily limited to one word, and may include multiple words or sentences. The search term extraction unit 131 extracts search terms from the keywords recorded in the search list using existing text analysis technology or the like. It should be noted that the search term is not limited to a word, and is made to correspond to a term registered in the dictionary DB 135, which will be described later.

The search term attribute assigning unit 132 performs a process of assigning an attribute to each search term extracted by the search term extracting unit 131 . Attributes include classification and destination flags, and the dictionary DB 135 is referred to when assigning attributes. The dictionary DB 135 is a database that records in advance information on classification and destination flags for each term. The dictionary DB 135 may be provided outside the destination estimation device 100 .

In the dictionary DB 135, the categories are hierarchically arranged in the order of, for example, “leisure”→“skiing”→“ski resort”. For example, if the term is about leisure in general, the category would be "Leisure", if the term was about skiing in general, the category would be "Skiing", which is a subcategory of "Leisure", and the term would be , the classification is "ski resort" which is a child classification of "ski". By hierarchically defining the classification, it is possible to grasp the terms of the parent classification and the terms of the child classification by referring to the dictionary DB 135 .

In addition, a plurality of classifications can be set for one term. For example, two classifications such as "skiing" and "climate" can be set as the classification of the term "snowfall information".

The destination flag is a flag that indicates whether the term can be a destination in order to determine whether the term is related to the destination. For example, the term "skis" may not be a destination, but the term "AA Ski Area" may be. Here, turning on the destination flag indicates that the term can be the destination.

The search term evaluation unit 133 evaluates the search term group to which attributes have been assigned by the search term attribute assigning unit 132, and performs processing to obtain the appearance frequency, which is an index indicating the appearance frequency for each classification, and the temporal appearance tendency. . Specifically, grouping is performed for each classification assigned to the search term, the frequency of appearance is counted for each classification, and the tendencies of appearance are classified based on the search date and time information attached to the search term. Appearance tendencies, for example, patterns such as steady appearance, intermittent appearance, appearance on weekends, appearance after holidays, and recent rapid increase, etc., and criteria for each are determined in advance, and one or more applicable patterns are determined. It can be classified by selection.

The inferred destination candidate extraction unit 134 performs a process of extracting, as an inferred destination candidate, a search term for which the destination flag is ON, among the search terms included in the classification with a high appearance frequency. It is possible to extract a plurality of estimated destination candidates. Candidates can be extracted.

The user characteristic evaluation unit 140 is a block that evaluates the characteristics of individual users, and includes a template matching unit 141 , a preference information processing unit 142 and a template DB 143 .

The template matching unit 141 performs a process of classifying the behavioral characteristics of the user based on the search word, the appearance frequency of the classification, the appearance tendency, and the position information. The user's behavioral characteristics are determined by referring to the template DB 143 in which models of user's behavioral characteristics are recorded together with criteria.

Models of behavioral characteristics include, for example, "Visit the place searched on the weekend the following weekend", "Visit the searched place the next day", "Visit the place when searching for a place that is not usually searched", and "Go out on Sundays". ``Search for a place to visit over a long period of time'', ``Visit the place when searching for a map'', ``Visit a place searched in the morning in the afternoon'', ``No relation between search and movement'', etc. For each model, judgment criteria based on the frequency of appearance of search terms and categories, appearance trends, and location information are defined.

The template matching unit 141 extracts the information about the search word attached with date and time information used by the estimated destination candidate extraction unit 130 in the process of extracting the estimated destination candidate, and the date and time information of the user acquired by the position information acquisition unit 120. Classification is performed by determining one or a plurality of types that apply to the behavioral characteristics of the user from the attached position information.

For example, if a search for a golf course is classified as appearing on the weekend, and the location information for that golf course is often obtained on the It will be classified as a behavioral characteristic of "Visit on weekends".

The preference information processing unit 142 processes preference information about user's preferences. For example, if the user has previously registered a preference target, that target is recorded as preference information. The object of preference can be, for example, “skiing”, “fishing”, etc., and it is desirable to correspond to the classification and terminology of the dictionary DB 135 .

In addition, the preference information processing unit 142 updates the preference information based on the information regarding the search words used by the inferred destination candidate extraction unit 130 in the process of extracting the inferred destination candidates. For example, when "baseball" is not included in the preferences registered by the user, and "baseball" is significantly increased in searches, "baseball" can be added to the preference information. Conversely, if "fishing" is recorded in the preference information and no search related to "fishing" has been performed for a long period of time, "fishing" may be deleted from the preference information.

The estimated destination determination unit 150 is a block that determines one or more estimated destinations from the estimated destination candidates extracted by the estimated destination candidate extraction unit 130. The estimated destination determination unit 150 includes an evaluation control unit 151 and a factor evaluation unit 152. there is

The evaluation control unit 151 performs processing for determining an inferred destination based on the appearance frequency of the inferred destination candidates in the search and the evaluation in consideration of various factors performed by the factor evaluation unit 152 . The factor evaluation unit 152 evaluates the estimated destination candidates with respect to factors such as user's preference, behavioral characteristics, and appearance tendency. However, the factors evaluated by factor evaluation unit 152 are examples, and are not limited to this example.

User preference factors are evaluated by referring to preference information updated by the preference information processing unit 142, and behavioral characteristic factors are evaluated by referring to user behavioral characteristic types classified by the template matching unit 141. In the evaluation of the tendency factor, the appearance tendency classified by the search term evaluation unit 133 is referred to.

The estimated destination presentation device 230 presents the estimated destination determined by the destination estimation device 100 to the user. Presentation of the estimated destination can be performed, for example, by displaying the estimated destination on the screen. When there is one guessed destination, an instruction as to whether or not to set the guessed destination as a route search destination is accepted, and when there are a plurality of guessed destinations, one of the guessed destinations is selected. It is desirable to receive an instruction as to whether to set the route search destination or not to set any of the estimated destinations as route search destinations.

The route search device 240 is a device that searches for a route to a destination set by the user. The route search device 240 can be composed of an in-vehicle navigator device, a service server that provides a route search service, or the like.

The above-described route search system 10 can be implemented with various equipment configurations. For example, as shown in FIG. 2( a ), a route search system 10 is configured by a PC, a smartphone, and an in-vehicle navigator connected to the Internet 310 . Functioning as the information acquisition device 220 , the in-vehicle navigator device can function as the destination estimation device 100 , the position information acquisition device 220 , the estimated destination presentation device 230 , and the route search device 240 .

In this case, the keyword searched by the user using a PC or smart phone is acquired for determining the estimated destination. In addition, the location to which the user has moved while carrying the smartphone and the location to which the vehicle equipped with the in-vehicle navigator device has been moved are acquired as location information. The in-vehicle navigator determines the estimated destination, and searches for a route to the destination in the in-vehicle navigator.

Further, as shown in FIG. 2(b), a route search system 10 is configured by a PC, a smartphone, an in-vehicle navigator device, and a destination estimation service server connected to the Internet 310, and the PC functions as a search device 210. functions as the search device 210 and the position information acquisition device 220, the in-vehicle navigator device functions as the position information acquisition device 220, the estimated destination presentation device 230, and the route search device 240, and the destination estimation service server functions as the destination estimation device 100 can function as

That is, a destination estimation service server that performs the destination estimation service is provided independently of the vehicle-mounted navigator. In this case, the destination estimation service server determines the estimated destination, transfers the estimated destination to the in-vehicle navigator, and performs route search. The keywords searched by the user on a PC or smartphone are acquired for determining the inferred destination, and the location where the user travels while carrying the smartphone and the location where the user travels by car equipped with an in-vehicle navigator device are acquired as location information. , is the same as the example shown in FIG.

Further, as shown in FIG. 3(a), a PC connected to the Internet 310, a smartphone, and an in-vehicle navigator device constitute a route search system 10, and the PC functions as a search device 210 and a destination estimation device 100, A smartphone can function as the search device 210 and the location information acquisition device 220 , and an in-vehicle navigator device can function as the location information acquisition device 220 , the estimated destination presentation device 230 , and the route search device 240 .

In this case, the PC determines the estimated destination, transfers the estimated destination to the in-vehicle navigator, and performs route search. The keywords searched by the user on a PC or smartphone are acquired for determining the inferred destination, and the location where the user travels while carrying the smartphone and the location where the user travels by car equipped with an in-vehicle navigator device are acquired as location information. , is the same as the example shown in FIG.

Further, as shown in FIG. 3B, a PC connected to the Internet 310, a smartphone, and an in-vehicle navigator device configure the route search system 10, the PC functions as a search device 210, and the smartphone serves as a destination estimation device. 100 , the search device 210 , the location information acquisition device 220 , and the estimated destination presentation device 230 , and the route search service server can function as the route search device 240 .

In this case, the smartphone determines the estimated destination, requests the route search service server to search for a route, and displays the searched route. The keywords searched by the user on a PC or smartphone are acquired for determining the inferred destination, and the location where the user travels while carrying the smartphone and the location where the user travels by car equipped with an in-vehicle navigator device are acquired as location information. , is the same as the example shown in FIG. Note that the device configurations shown in FIGS. 2 and 3 are examples, and the route search system 10 may be realized with other device configurations.

Next, the operation of the destination estimation device 100 that constitutes the route search system 10 will be described. The operation of the destination estimation device 100 can be divided into an operation of extracting estimated destination candidates and an operation of determining an estimated destination. The operation of extracting inferred destination candidates is an operation of extracting inferred destination candidates from the obtained search list. The speculative destination determination operation is an operation of determining an inferred destination from among the inferred destination candidates in consideration of various factors. It should be noted that the user behavior characteristic template matching operation performed by the user characteristic evaluation unit 140 and the processing of the user's preference information are appropriately performed according to the above-described criteria.

In the operation example described below, the candidates for the estimated destination are extracted in advance, and the estimated destination is determined from among the estimated destination candidates when setting the destination. As a result, it is possible to save the trouble of extracting an estimated destination for each destination setting, and to determine an estimated destination suitable for the date and time of setting the destination. However, the timing for extracting candidates for the estimated destination and the timing for determining the estimated destination can be determined arbitrarily.

First, the operation of extracting estimated destination candidates will be described with reference to the flowchart of FIG. Since it is necessary to accumulate a certain amount of keywords in the search list, the operation of extracting the estimated destination candidates may be performed at a predetermined timing (S101). The predetermined timing can be, for example, periodically or when the size of the search list reaches a predetermined amount.

If it is time to perform an operation of extracting an inferred destination candidate (S101: Yes), the search list acquisition unit 110 acquires a search list from the search list creation unit 212 of the search device 210 (S102).

Then, the search term extraction unit 131 of the inferred destination candidate extraction unit 130 extracts search terms from the keywords recorded in the search list (S103). As described above, search term extraction can be performed using existing text analysis techniques.

When the search term is extracted, the search term attribute assigning unit 132 refers to the dictionary DB 135 and assigns an attribute to each search term (S104). Attributes assigned to search terms include classification and destination flags. A plurality of classifications may be assigned to one search term.

Next, the search term evaluation unit 133 aggregates the appearance frequency for each search term classification, and classifies the appearance tendency based on the search date and time attached to the search term (S105). In summing the frequency of appearance, grouping is performed for each classification of search words.

Since the classification is hierarchical, for example, grouping as shown in FIG. 5 can be performed. In this figure, "AA ski resort", "BB ski resort", "skis", "rental ski", "snow cover information", "XX baseball field", "CC ramen", "DD town lunch", and "ramen" were extracted as search terms. Take the case as an example. It should be noted that a search word surrounded by a square indicates that the destination flag is on, that is, that it can be the destination.

Here, "AA ski resort" and "BB ski resort" are classified as "ski resort", "skis" and "rental skis" are classified as "ski equipment", and "snow coverage information" is classified as "ski resort". and "Climate", "XX Baseball Field" is classified as "Baseball Field", "CC Ramen" and "DD Town Lunch" are classified as "Restaurant", and "Ramen" shall be classified as “eating and drinking”.

In the dictionary DB 153, "ski" is set as the parent classification of the classification "ski resort" and "ski equipment", "baseball" is set as the parent classification of the classification "baseball field", and "ski" and "baseball" are set as the parent classification of It is assumed that "leisure" is set and "restaurant" is set as the parent classification of the classification "restaurant". The inclusion relationship of the categories shown in the figure is based on this parent-child relationship.

The counting of appearance frequencies can be performed, for example, by paying attention to the lowest layer classification. At this time, it is assumed that the appearance frequency of the search term included in the parent classification is inherited to the child classification. For example, the frequency of appearance of the search term for “ski resort”, which is the lowest class of classified leisure, is 2 (AA ski resort, BB ski resort), but the number of search terms for the parent classification “ski” is 2 (AA ski resort, BB ski resort). It is possible to add 1 appearance frequency (snowfall information) to 3. As for the classification "climate", since it is the lowest layer, the appearance frequency for "climate" is 1. By performing similar aggregation for each category, the aggregation frequency for each category can be aggregated.

Also, by referring to the search date and time information of the search word, it is possible to determine the appearance tendency such as steady appearance, intermittent appearance, appearance on weekends, appearance after holidays, recent rapid increase, etc. for each classification.

Finally, the inferred destination candidate extraction unit 134 extracts search terms whose destination flags are turned on, among the search terms included in the classification with the highest appearance frequency (S106). The extracted search term is extracted as an inferred destination candidate.

In the example shown in FIG. 5, "AA ski resort", "BB ski resort", and "CC Ramen" included in the classification with the appearance frequency of 3 are extracted as the estimated destination candidates. "Skis" and "Rental skis" are included in the classification with the appearance frequency of 3, but are not extracted because the destination flag is off.

Next, the operation of determining the estimated destination will be described with reference to the flowchart of FIG. The operation of determining the estimated destination is preferably performed when the estimated destination presentation device 230 is activated. For example, when the in-vehicle navigator device functions as the presumed destination presentation device 230, the presumed destination is determined by including the information at that point in the factor evaluation when the engine of the automobile is started. Therefore, when the user turns on the engine, the guessed destination will be automatically presented.

In addition, when the smartphone functions as the estimated destination presentation device 230, the estimated destination is automatically presented when the application for route search is launched. In either case, the user can set the destination of the route search without being aware of the task of inputting the destination.

In the operation of determining an inferred destination, the factor evaluation unit 152 performs an evaluation considering various factors for each of the inferred destination candidates with a high frequency of occurrence, and the evaluation control unit 151 determines the inferred destination with a high overall evaluation. A location candidate is determined for speculation purposes. In this example, the evaluation is performed considering the preference factor, the behavioral characteristic factor, and the appearance tendency factor, but other factor evaluation may be adopted.

First, the estimated destination candidates obtained by the above-described estimated destination extraction operation are acquired (S201). Also, current information is acquired (S202). The current information can include the time of day, day of the week, and the like. Depending on the factor evaluation employed, information such as temperature and weather may also be included.

In the preference factor evaluation (S203), the user's preference information updated by the preference information processing unit 142 is referred to, and points are added to the estimated destination candidates that match the preference information. Whether or not the destination matches the preference information can be determined by referring to the classification including the parent classifier classification of the inferred destination candidate.

In behavioral characteristic factor evaluation (S204), points are added to inferred destination candidates that match the user's behavioral characteristics classified by the template matching unit 141. FIG. For example, if the behavioral characteristic of the user is "Visit the place searched on the weekend next weekend", the search frequency for the speculated destination candidate was high last weekend, and if the current information is the weekend, then the behavioral characteristic Points are added to the inferred destination candidate as a match. In this case, if the current information is a weekday, no points will be added.

In the occurrence tendency factor evaluation (S205), points are added to inferred destination candidates whose occurrence tendency indicates characteristics that are assumed to be highly likely to be set as the destination. For example, points are added to inferred destination candidates classified as "constantly appearing" or "rapidly increasing recently".

The evaluation control unit 151 comprehensively evaluates the appearance frequency and each factor evaluation (S206), and determines an estimated destination (S207). For example, the points of each factor evaluation are added to the points corresponding to the appearance frequency, and the guessed destination candidate with the highest total points is determined as the guessed destination. At this time, a plurality of estimated destination candidates may be determined as the estimated destination in descending order of total points. Of course, various methods can be used for the comprehensive evaluation of the appearance frequency and each factor evaluation.

For example, if the user has recently searched for "AA ski resort" among "AA ski resort", "BB ski resort", and "CC ramen" extracted as inferred destination candidates, the appearance tendency factor evaluation Since points are added, if other factor evaluations are equal, "AA Ski Resort" will be determined as the guessed destination.

In addition, if the user's behavioral characteristics are classified into, for example, "going for leisure on weekends" and "eating out on weekdays", if the destination setting date is a weekday, the behavioral characteristics factor evaluation of "CC ramen" is performed. becomes high, and if the destination setting date falls on a weekend, the behavioral characteristic factor evaluation of "AA ski resort" and "BB ski resort" becomes high.

As described above, the route search system 10 of the present embodiment extracts inferred destination candidates based on search terms used by the user not for route searches but for daily searches. Therefore, there is no need to be conscious of the task of inputting the destination. In addition, since the estimated destination is determined from among the extracted estimated destination candidates based on the information at the time of the route search and the behavioral characteristics of the user, the user can search the route without being conscious of the task of inputting the destination. You will be able to set the destination of

It should be noted that the estimated destination candidates extracted by the operation of extracting the estimated destination candidates may be presented to the user. Even in this case, the user only has to select a destination from among the candidates, so the user is freed from the task of inputting the destination.

10... Route search system 100... Destination guessing device 110... Search list obtaining unit 120... Position information obtaining unit 130... Guessed destination candidate extracting unit 131... Search word extracting unit 132... Search word attribute assigning unit 133... Search word evaluating unit 134 Estimated destination candidate extraction unit 135 Dictionary DB
140 User characteristic evaluation unit 141 Template matching unit 142 Preference information processing unit 143 Template DB
150: Guessed destination determination unit 151: Evaluation control unit 152: Factor evaluation unit

Claims (1)

a search information acquisition unit that acquires search information including information about a search term searched on a search site;
an attribute adding unit that attaches attribute information indicating an attribute of the search term to the search term;
a search term evaluation unit that obtains, for each attribute, the appearance frequency of search terms that share the same attribute;
an inferred destination candidate extraction unit that extracts, as an inferred destination candidate, a search term related to a destination from among search terms included in attributes whose frequency of appearance satisfies a predetermined criterion;
A destination estimation device comprising:

Images