JP2020086740A - Store search method, server and program - Google Patents

Abstract

To further thoroughly search for restaurants that offer the desired foods and drinks.SOLUTION: A food search server 101 comprises: a search request reception module 201 for accepting access from a search terminal 111, receiving a food name that is a search key to store and manage the same in a database; a food name fluctuation processing module 202 for absorbing fluctuation in a food name of the food indicated by the received food name to identify the food; a store list generation module 203 for outputting, as a list, stores obtained as a result of the search, such as restaurants that provide desired foods; and a database management module 204 for creating and updating, etc. various databases. The search terminal 111 comprises: a screen control module 205 that displays an input screen for searching and displays a search result list received from the server; and a search result processing module 207 that displays the search results received from a data transmission and reception module 206 and the server in a predetermined format.SELECTED DRAWING: Figure 2

Description

The present invention relates to a store search method, a server, and a program, and more specifically, to a store search method, a server, and a program that can perform various searches for a store or a facility that mainly or subordinately provides food and drink.

Online search systems are widely available to search for restaurants and facilities such as inns and hotels that provide food, but usually the location of the store, the genre or price range of the store such as Japanese or Chinese, etc. The search is performed using the key, and there are cases where the store that provides the desired food and drink cannot be searched. That is, when it is not possible to directly search using the food and drink name as a key, for example, the search will be performed using the genre of the restaurant as a key, but all the stores of the supposed genre provide the desired food and drink. Not necessarily, and conversely, a store of a genre other than the expected genre may provide desired food and drink.

As such a search system, for example, a user who is unfamiliar with the land to be searched and a restaurant that suits the taste of each user residing on the land are recommended for users who search for the restaurant. The information providing system comprises a server device 70 that searches the database 72 based on the search information from the terminal device 10 and returns the searched result to the information terminal device 10 for the purpose of Proposes an information providing system characterized by determining a search method for the database 72 based on the storage location associated with the information terminal device 10 and the current position information of the information terminal device 10 (Patent Document 1). See).
JP, 2009-245275, A

However, since the conventional store search system is a search system that uses the genre of the store as a key, it is difficult to search for a desired dish. In addition, even if you can search on the menu, it is normal for food and drinks that are generally offered at restaurants to have slightly different product names (fluctuating food and drink names), such as food and drink names for the same Chinese food. However, since the notation and name are different depending on the origin, region, school, etc., if you simply search by the food name provided by the store, only the store that uses the food name that matches the food name There is a problem in that even if there is a famous store among the stores that have not been searched and other notations are used, the search cannot be performed.

The present invention has been made in view of the above-mentioned conventional problems, and stores that use different notation food/drink names if the food/drink name fluctuations of food/drinks provided by the store are appropriately processed and the same food/drink are searched. Therefore, it is an object of the present invention to provide a store search method, a server, and a program that can search for restaurants that provide desired food and drink with less leakage.

The invention according to claim 1 is a food and drink search system, and which food and drink is processed by receiving means for receiving a desired food and drink name and food and drink name fluctuations included in the received food and drink. For all food and drink names generated by the food and drink name fluctuations that specify the food and drink specifying means that specifies the specified food and drink, the stores that provide food and drink with the food and drink name are searched from the product name database to provide the food and drink. And a list generation means for generating a store list.

According to the invention of claim 2, in the food and drink search system according to claim 1, the food and drink specifying means has a food and drink name fluctuation dictionary, and refers to the food and drink name fluctuation dictionary to specify food and drink. Characterize.

The invention according to claim 3 is the food and drink product search system according to claim 2, wherein the food and drink name fluctuation dictionary stores the food and drink name used as a name indicating the food and drink for each food and drink in association with each other. The feature is that the food and drink can be identified from the food and drink name.

The invention according to claim 4 is the food and drink search system according to claim 1, characterized in that the food and drink specifying means specifies the food and drink from the desired food and drink name by machine learning.

The invention according to claim 5 is, in the food and drink retrieval system according to any one of claims 1 to 4, a representative database that stores food and drink names of representative dishes that are popular dishes for customers who represent each store, For all food and drink names that occur due to fluctuations in food and drink names of food and drink products that represent the specified food and drink products, search the stores that provide food and drink products with the food and drink product names from the representative database, and list the representative food stores that use the specified food and drink products as representative food And a representative list generating means for generating.

The invention according to claim 6 is the food/drink retrieval system according to claim 5, wherein the representative database collects comment information for the stores posted via the network, and the frequently-appearing food/drink names are representative of the store. It is characterized by being generated by storing it as a dish in association with a store.

According to the present invention, a receiving unit that receives a desired food and drink name, a food and drink specifying unit that processes the food and drink name fluctuations included in the received food and drink and specifies which food and drink corresponds to the specified food and drink, For all food and drink names generated by the food and drink name fluctuation of the food and drink name indicating the food and drink, the store that provides the food and drink of the food and drink name is searched from the product name database, and a list generating means that generates a food and drink provision store list is provided. Therefore, by using the food and drink name that the searcher desires to search, the same food and drink can be searched for with less omissions, although it is described by other food and drink names.

FIG. 1 is an overall system configuration diagram of an embodiment of the present invention. It is a functional block diagram of an example of a food search server and a search terminal of one embodiment of the present invention. It is a figure for demonstrating food and drink name fluctuation of one embodiment of the present invention. It is a flow chart which shows processing in a food and drink search server of one embodiment of the present invention. It is a figure which shows the search screen of the search terminal of one Embodiment of this invention. It is a flow chart which shows processing which generates a typical database of one embodiment of the present invention.

Embodiments of the store search method, server, and program of the present invention will be described below with reference to the drawings. Note that the same reference numerals are used to indicate the same processing and configuration in different drawings.

An object of the present invention is to search for a store that provides such food and drink appropriately and with less omission by using a desired food and drink name as a key. That is, in the conventional restaurant search system with the genre or region as the key, for example, even if an attempt is made to search for a store offering water-cooking, a store specializing in water-cooking can be searched, but a pub serving water-cooking cannot be searched. In addition, even if you try to search for a restaurant that offers kaibein, water dumplings, or fin cutlets, you can only search for Chinese food or a tonkatsu restaurant.Therefore, you should inquire directly to the store as to whether you serve kaibein, water dumplings, or fin cutlets. Absent. Furthermore, even if you can search for Japanese food and Chinese food as well, not all Japanese restaurants and Chinese restaurants provide grilled fish and levanilla, so search for a restaurant that offers these menus. I can't. In addition to this, there are many requests to search for accommodation facilities such as hotels and inns that provide specialty dishes and other commercial facilities, and it is also an object to realize this.

Here, such a search seems to be feasible by using a product name database that associates food and drink names with the stores that provide them, but it is easily realized due to the peculiarity of handling food and drink product names. It is not possible. Generally, foods and drinks provided in restaurants and the like have slightly different product names even if they are almost the same food and drink. For example, even if the food and drink name of the same Chinese food is different in the notation and name depending on the origin, region, school etc., if you simply search only the food and drink name provided by the store, the food and drink that matches the food and drink name Only stores that use the product name can be searched, and even if there are prominent stores that use other notations, you cannot search. Specifically, referring to FIG. 3, even if it is called the same ramen, there are different notations such as Chinese noodles, ramen, char siu noodles, etc. In addition, although it is not exactly the same, generally when searching as ramen It is necessary to consider the concept of foods and drinks that are supposed to be searched (hereinafter referred to as similar foods and drinks names) including tanmen and tsukemen (hereinafter referred to as food and drinks name fluctuations). In the present embodiment, when searching for a desired food or drink, such a fluctuation in food or drink name is absorbed, and a suitable food or drink is identified and searched as a food or drink corresponding to the food or drink name input by the user who searches. With this, it becomes possible to perform a search with less omission.

(First embodiment)
First, a system for performing a search using a food name as a key will be described as the first embodiment. FIG. 1 is an overall system configuration diagram of one embodiment of the present invention, and FIG. 2 is a functional block diagram of an example of a food search server and a search terminal of one embodiment of the present invention. In addition, FIG. 3 is a diagram for explaining the fluctuation of the food and drink name according to the embodiment of the present invention, and is a diagram showing an example of associating the food with the food name. In addition, although the search key is food in the present embodiment, the search key is not limited to this and may be beverages and foods and drinks including beverages.

(System configuration)
In the present embodiment, as shown in FIG. 1, the food search server 101 that performs processing necessary for the store search system such as transmission/reception with each terminal, data management, search processing, database construction/management, and posting of comments. A comment posting server 102 that receives and provides posted comments on a site or the like is connected to a network, and is connected to a user-side search terminal 111 or general terminal 112 via the network 103. The comment posting server 102 may be any server as long as it provides a service of posting and writing comments, and is a server providing, for example, SNS, restaurant search service, blog, etc. The server can be any server that provides a service in which a user posts or writes information or impressions about a restaurant.

In the present embodiment, the functions of the food search server 101 and the comment posting server 102 can be provided to each terminal through the cloud. In this case, each terminal mainly uses a browser or a dedicated application to perform only input and display processing, and data management and various processing are basically performed on the server. The configuration may be such that it is directly provided to a client or a terminal via a dedicated network.

In the present embodiment, the search terminal 111 and the general terminal 112 are used as the terminals on the user side. The search terminal 111 and the general terminal 112 are connected to the food search server 101 for the user to search, and the customer who visits the store comments. It connects to the posting server 102 and posts a comment. Although the user of the search terminal 111 may be a visitor to the store, there are usually a considerable number of visitors, and many comments are posted and posted from many general terminals 112.

The terminals 111 and 112 are basically wirelessly connected to the network 103 and can be, for example, tablet terminals. In the present embodiment, the connection to the network 103 is performed by a mobile phone line, Wi-fi, or BLUETOOTH. (Registered trademark). Here, the terminals 111 and 112 can be not only tablet terminals but also smartphones or mobile personal computers, but basically, each server is accessed via a network to display an input screen, and a touch panel, a mouse or a keyboard is used. Any device such as a mobile or desktop personal computer or a dedicated terminal can be used as long as a certain input operation can be performed.

Further, as the server of the present embodiment, the processing is performed only by the food search server 101, but the present invention is not limited to this, and it is also possible to divide the functions into a plurality of servers and share the functions.

FIG. 2 is a functional block diagram of the food search server and the search terminal according to the embodiment of the present invention. The food search server 101 is accessed from the search terminal 111, receives a food name serving as a search key, stores it in a database, and manages it. A search request reception module 201, food name fluctuation of food indicated by the received food name. Is absorbed and the food name fluctuation processing module 202 that identifies the food, the store that provides the identified food is searched, and the stores such as the restaurants that provide the desired food obtained as a result of the search are output as a list. It includes a store list generation module 203 for operating and a database management module 204 for generating and updating various databases.

On the other hand, the search terminal 111 displays an input screen for searching, displays a search result list received from the server, sends a search key to the server, and sends a search key to the server. A data transmission/reception module 206 for receiving display data and the like, and a search result processing module 207 for processing the search result received from the server into a predetermined format and displaying it.

(Food and beverage name fluctuation processing)
With the above-described configuration, the search is performed using the food name as a key. In the present embodiment, the association of food names as shown in FIG. 3 is managed in a dictionary format (not shown), and such a dictionary is used. By appropriately absorbing the food name fluctuations, the food is specified, the restaurant that provides such food is placed in a database (not shown) to be searched, and the corresponding restaurant is extracted.

First, the food and drink name fluctuation will be described. As shown in FIG. 3, the ramen 301 which is a food includes not only ramen as a food name but also different notations such as Chinese noodles, ramen, and char siu noodles. Further, the names of similar foods and drinks include tanmen, tsukemen, etc., which are not exactly the same but are generally assumed to be searched when searching for ramen, but are not limited thereto, and are known in the technical field. Can include food name notation. On the other hand, the food soba 302 includes buckwheat noodles and tsukemen in addition to soba as a food and drink name, and tsukemen is also a food and drink name of the same kind of ramen as food. As described above, the food-food name association of the present embodiment has a structure in which the food name of one food may be included in the food names of other foods. Furthermore, the same noodles as ramen 301 and soba 302 can be made to include udon 303, and Chinese noodles and yakisoba, etc. can be included as Chinese noodles other than ramen 301, and fried rice can be included in addition to noodles. You can also This is to realize a more comprehensive search with less omission by capturing a desired food more broadly, but the invention is not limited to this. Any method known in the art can be used to adopt a taxonomy that appropriately processes food and drink name fluctuations. In the present embodiment, the upper classifications such as noodles and Chinese are simply described for convenience and not particularly used, and the ramen is not included in Chinese, but of course, it can be included in Chinese.

Given the classification system as described above, the processing of the food and drink name fluctuation of this embodiment will be described with reference to FIGS. 4 and 5. FIG. 4 shows the processing in the food and drink search server of one embodiment of the present invention. 5 is a flowchart, and FIG. 5 is a diagram showing a search screen of the search terminal according to the embodiment of the present invention. When using ramen as the food name to be searched, the user first enters ramen into the food name 502 on the search screen 501 of the search terminal 111 and presses the search button, and a search request for the food name "ramen" is sent to the food search server 101. The search request receiving module 201 receives the request (step 401). The food name fluctuation processing module 202 processes the food name fluctuation from the received food name to identify the food (step 402). In the present embodiment, the ramen 301 is specified as the food by referring to the food and drink name fluctuation dictionary (not shown) generated from the association shown in FIG. The food ramen 301 includes ramen, Chinese soba noodles, ramen, tsukemen, tanmen, and char siu men as food names, so the store list generation module 203 refers to the product name database for any one of the food names and is specified. The restaurants that provide food with the specified food name are extracted (step 403).

Here, the product name database is stored by associating the food and drink names with the stores that provide them, and by extracting the stores to be provided by using the food and drink names as keys, it is also possible to obtain the store information. Any database known in the art can be used if present. The product name database can collect information from each store by, for example, sending a questionnaire to each restaurant, or can be configured in advance by collecting the information by crawling the Internet. That is, a menu image of a restaurant is collected on the Internet, characters are extracted from the collected menu image and converted into character information, and a menu providing a menu is provided from the converted character information or information accompanying the menu image. It can also be created by acquiring the store name and other information, extracting the product name of the dish, and associating it with the restaurant.

In the case of the “ramen” of the present embodiment, since there are a plurality of food names, it is determined whether or not the search process for one food name has been executed for all food names (step 404), and all food names If the search has not been executed (No), the search is continued. If all have been executed (Yes), the extracted restaurants are merged to generate a list of stores that provide the searched food (step 405). The generated list is sent to the search terminal, and the search result display module 207 displays the list of restaurants on the search result screen (not shown).

As described above, according to the present embodiment, by using the food and drink name that the searcher desires to search, foods that are considered to be the same food and drink can be represented by other food and drink names despite the fluctuation of the food and drink name. You can search without omission.

(Second embodiment)
In the above-described first embodiment, the food name is used as a key to search based on the product name database in which the food and drink provided by each restaurant and the like are associated with each other. That is, basically, since the search is performed using a product name database that includes food and drinks provided by the restaurant, it is possible to extract many restaurants that provide desired food and drink with high comprehensiveness. However, since the number of stores to be extracted increases, it is necessary to further narrow down the conditions or to make a final selection by referring to the detailed information of each restaurant. In the present embodiment, as a database for searching restaurants, a representative database that represents each restaurant and that includes only one or at most several foods and drinks of a representative dish that is popular with customers is used.

That is, since the food is searched based on the representative database that associates one or several representative dishes with each restaurant for each restaurant, not only the desired food but also the representative dish of the restaurant is extracted. Therefore, for example, it is effective as a reference for selecting a restaurant desired to visit. Further, as compared with the search of the first embodiment, it is possible to considerably narrow down the search only by the food and drink name. Here, although it is a representative dish of each restaurant, when the objective selection is made including the quality of the dish, the representative dish may be 0 depending on the restaurant, so it is considered to be more effective.

(Generation of representative database)
In the present embodiment, as an example of the method of generating the representative database, the representative database is generated based on various comments posted on the net, but the present invention is not limited to this. It can be generated by any method known in the art, such as obtaining information and generating a representative database.

In recent years, a user who visits a restaurant is more likely to post impressions of the visit and information about the restaurant (hereinafter, collectively referred to as a comment) on the SNS, a restaurant search site, a blog, and the like, and to write the information. And is a valuable source of information when choosing a restaurant. For example, the comment is posted by accessing the comment posting server 102 shown in FIG. 1 from the general terminal 112. Specifically, when the comment posting server 102 is an SNS server, the user connects to the comment posting server 102 from the general terminal 112 and posts the comment on his or her own timeline, so that the comment is posted and posted on the net. Searchable.

In the present embodiment, the representative database is generated using the comments that are searchable in this way, by the processing as shown in FIG. FIG. 6 is a flowchart showing a process of generating a representative database according to the embodiment of the present invention. Every certain period of time, etc., the site on the Internet is crawled to search for comments on restaurants (step 601). The food and drink names are extracted from the comments for the searched store (step 602), and the process is performed until the extraction of all comments for the searched store is completed (step 603). The extracted food and drink names are arranged in descending order of the number of appearances, and the food and drink names of a predetermined rank, for example, up to the third rank are stored as a representative dish in the representative database in association with the restaurant (step 604). The above processing is executed for all the restaurants included in the database such as the product name database (step 605). Regarding the determination of the representative food, in addition to determining up to a predetermined rank as the representative food, if the number of appearances of the first place and the second place is significantly different, only the food and drink name of the first place of appearance may be adopted. It is possible to use any method of appropriately determining the representative dish based on the comment regardless of the number of appearances.

As a result, the number of stores to be extracted increases, and it is not necessary to narrow down by other conditions or finally select by referring to the detailed information of each restaurant, and a restaurant that provides desired food and drink. You can search for an accurate restaurant.

(Third Embodiment)
In the first and second embodiments described above, the food and drink name fluctuation is processed by referring to the food and drink name fluctuation dictionary to enable a more appropriate search, but in the present embodiment, the food and drink name fluctuation is processed as follows. This is done by machine learning. That is, the feature amount related to the food and drink name of each food and drink is obtained by inputting and learning a large number of food and drink name information, and it is possible to specify the food and drink even if there is fluctuation in the food and drink name. By using the food/drink name desired to be searched, the same food/drink can be searched more thoroughly, although it is described by other food/drink names.

Specifically, basically, in the search process of the first and second embodiments, the same process is performed in the present embodiment except for the process of identifying the food item from the received food name (step 402). The processing of can be executed.

(Fourth Embodiment)
In the present embodiment, in addition to the functions provided in the above-described first to third embodiments, a function to enable registration of a hidden list so as not to display a store including a specific key from search results I will provide a. Therefore, in the present embodiment, personal information can be registered for each user, and the hidden key is not limited to information that can be basically a search key such as a food/drink name, region, and genre of the store. Various information that may be included in the information can be included. Further, since a key to be hidden can be set depending on the time zone and the position information, it is possible not to display the store that provides only confectionery such as manju during the dinner time zone. Further, it is possible to display only the shops of the genre used for entertaining or working with colleagues in weekday search, and not display such shops on weekends.

As a specific process, when displaying the stores searched in the first to third embodiments described above, the personal settings of the user are referred to, and the stores are hidden in the stores corresponding to the hidden items depending on the time zone and the position information. The flag is set so that it is not displayed when the list is displayed, but the present invention is not limited to this, and it can be hidden by a method other than the flag known in the technical field. The personal setting information can be set in the terminal or can be managed on the server.

Further, although the databases used in the above-described first to third embodiments have been described as being intended for restaurants, the present invention is not limited to this, and it is also possible to provide accommodation facilities such as inns and hotels, and search items. Can also include items specific to that facility. In other words, you can search for hotels that specialize in steak dishes, or you can search for hot spring inns that are famous for their prime, using the prime and hot springs as keys.

As described above, according to this embodiment, it is possible to search not only restaurants but also various facilities that provide food and drink according to various personal tastes.
[Other Embodiment 1] A receiving step of receiving a desired food and drink name,
A food and drink identifying step of processing the food and drink name fluctuations included in the received food and drink to identify which food and drink corresponds to the food and drink;
For all food and drink names generated by the food and drink name fluctuation of the food and drink name indicating the specified food and drink, the store that provides the food and drink of the food and drink name is searched from the product name database, and the food and drink providing store list is generated. A method for searching food and drink, comprising:
[Other Embodiment 2] A program for causing a computer to execute a restaurant search method, the method comprising:
A receiving step of receiving a desired food and drink name,
A food and drink identifying step of processing the food and drink name fluctuations included in the received food and drink to identify which food and drink it corresponds to,
For all food and drink names caused by fluctuations in food and drink names of food and drink names indicating the specified food and drink, a store that provides food and drink with the food and drink name is searched from the product name database to generate a food and drink providing store list A program comprising steps and.

Claims (6)

A food and beverage search system,
Receiving means for receiving the desired food and drink name,
A food and drink identifying means that processes the food and drink name fluctuations included in the received food and drink to specify which food and drink corresponds to the food and drink,
For all food and drink names generated by the food and drink name fluctuation of the food and drink name indicating the specified food and drink, the store that provides the food and drink of the food and drink name is searched from the product name database, and the food and drink providing store list is generated. A food and beverage search system, characterized by comprising means. The food-drinks search system according to claim 1, wherein the food-drinks identifying unit has a food-drinks fluctuation fluctuation dictionary, and identifies the foods-drinks by referring to the food-drinks fluctuation fluctuation dictionary. The food and drink name fluctuation dictionary stores the food and drink names used as names indicating the food and drink in association with each other, and the food and drink can be specified from the food and drink name. Food and beverage search system described in. The food and drink search system according to claim 1, wherein the food and drink specifying means specifies a food and drink from the desired food and drink name by machine learning. A representative database that stores food and drink names of representative dishes that are popular with customers representing each store,
For all food and drink names generated by fluctuation of food and drink names of food and drink names indicating predetermined food and drink, a store that provides food and drink with the food and drink name concerned is searched from the representative database, and the predetermined food and drink is a representative dish. A representative list generating means for generating a restaurant list is further provided. The representative database is generated by collecting comment information about a store posted via a network and storing frequently-appearing food and drink names in association with the store as a representative dish of the store. Item 5. A food/drink retrieval system according to item 5.

Images