JP6378514B2 - Financial information retrieval system - Google Patents

Description

  The present invention relates to a search technique, and more particularly to a technique that is effective when applied to a financial information search system that searches for information related to financial products.

  For example, in a financial institution such as a securities company, when a sales representative conducts sales activities for a customer, the topic or keyword that the customer is interested in or likely to have (hereinafter these are collectively referred to as “keyword”). Search) based on keywords related to matters that are generally discussed in the world, and materials for conversation and suggestions are obtained. Such operations are performed as preparations before conducting sales activities for customers, and may be performed at any time using a tablet terminal or the like according to the contents of conversations with customers at the site of sales activities. .

  Generally, for example, a keyword such as “sunlight” is specified on the screen, and information is obtained from the search results obtained through the search engine. In addition to targeting information on the Internet using an engine, it is also possible to provide high-value-added sales by targeting specialized information such as securities analyst reports held by financial institutions on their internal systems. It is done as a material of activity.

  In this case, for example, it has been easy and easy to extract a report that includes the word “sunlight”, but in actual sales activities, sales representatives read the extracted report, Search for information related to individual companies (brands) related to this, search for the brand information again, obtain stock charts and various related information related to the brand, and present them to customers. Approach is taken. That is, in the search based on the keyword, it is possible to obtain a related report, but it is rare that information relating to individual companies (brands) can be directly obtained, which places a heavy burden on the sales staff.

  As a technology related to searching for brand information, for example, Japanese Patent Application Laid-Open No. 2009-122806 (Patent Document 1) stores document data collected from a publicly traded company Web server in association with a brand code of each company. The company HP document DB, a keyword extraction unit that extracts a plurality of keywords from each document data, and stores each keyword and each appearance frequency in the keyword DB in association with the company code, and a search term is input If it is, the keyword DB is referred to, the brand code associated with the same keyword as the search word and the appearance frequency of each keyword are specified, and the search result data including the appearance frequency and the brand code is arranged in the order of appearance frequency. A search system including a search processing unit that transmits the list to the user's PC terminal is described.

JP 2009-122806 A

  When searching for information on stocks of financial products, regardless of what search method is used, the search results can be displayed in a state where the stocks are sorted according to a predetermined condition such as “order of dividend yield”. It is desirable to do so. Here, for example, when Google (registered trademark) or the like is used as a search engine, only the upper predetermined number (for example, 1000) of search results is output in consideration of the efficiency and speed of the search process. It is restricted to be a target, and this output result is sorted.

  As a result, depending on the sorting conditions, a brand that should originally be in a higher rank (for example, a brand with a high “dividend yield”) may be excluded from the sort target in the first place. Here, if the search conditions are further narrowed down for the search result (for example, further narrow down to “stocks that analysts say Buy”), the search results so far are reset once, and all brands are targeted again. The search is performed under the conditions narrowed down to “3”, and the search result is sorted again in the “dividend yield order”.

  At this time, if the number of search results that match the filtering conditions is less than the above-mentioned predetermined number, there are issues that were not included in the upper predetermined number before adding the filtering condition ( For example, a brand having a high “dividend yield” may be newly included. Accordingly, the sorted search result also includes the brand, and as a result, a brand that was not displayed before narrowing down the conditions may be suddenly displayed at the top of the search result. .

  In general search, it is common sense that the search results are sequentially narrowed down by adding and narrowing down the search conditions. Inconsistent results may occur in that brands that have not yet appeared as search results are suddenly displayed at the top of the search results. Such a situation is unlikely to be a major problem in a general search system, but the search target is a financial product, and sales representatives perform sales activities such as proposals for financial products based on the search results. If it is used, it may lead to violations of laws and regulations such as misleading solicitation, which is a big problem.

  Such a situation can occur regardless of whether or not a keyword is specified for the initial search. However, when searching by specifying a keyword, when extracting a predetermined number of high-order items, extraction is performed based on the matching rate of the keyword, whereas if no keyword is specified, the Web Extraction is performed based on another index such as the reference frequency of pages and data (for example, page rank (registered trademark) in Google (registered trademark)). Therefore, even if the above situation occurs, if a keyword is specified, it is assumed that, among various conditions, the result is sorted with priority on the match rate with the specified keyword. By explaining the sales person, it is possible to take care not to invite misidentification.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to display data that can be included in a predetermined upper number of items under a sort condition for a search result without omission even when a keyword is not specified during a search by a search engine. It is to provide a financial information retrieval system.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows.

  A financial information search system according to a representative embodiment of the present invention is a financial information search system in which a search engine searches for information related to brands of financial products via a network using an information processing terminal. , A crawler that creates a full-text search index table by crawling a set of document data described for each stock held on a data source at a predetermined timing, and a search request received from the information processing terminal. On the other hand, a search processing unit that responds as a search result of the full-text search among the records related to the document data that match from the index table is returned.

  In addition, when a keyword is not specified in the search request received from the information processing terminal, a search process is performed directly on the set of document data on the data source instead of a search by the search engine. Has a DB search unit.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  That is, according to the representative embodiment of the present invention, even when no keyword is specified when searching for a brand by a search engine, brands that can be included in a predetermined number of higher ranks under the sort condition for the search result. Can be displayed without leaking.

It is the figure which showed the outline | summary about the structural example of the financial information search system which is one embodiment of this invention. It is the figure which showed the outline | summary about the example of the data structure of the stock brand index in one embodiment of this invention. (A), (b) is the figure which showed the outline | summary about the example of the data structure of the document index in one embodiment of this invention. It is the figure shown about the example of the search screen of the brand information in the sales terminal in one embodiment of this invention. It is the figure which showed the outline | summary about the example in the case of sorting and outputting the search result by narrowing down in one embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

  According to the prior art, for example, it is possible to directly present a brand closely related to a keyword obtained from a conversation with a customer as an investment target. However, in the prior art, keywords are extracted in advance from a publicly traded company Web server, and there is no difference in information source and search using a search engine on a general Internet, such as an analyst report. It is not possible to target documents that can be referenced only in the in-house system. Even when an analyst report or the like is used as an information source, as described above, it is difficult to directly obtain information on individual companies (brands) related to the keyword, not a report including the specified keyword. It is.

  Therefore, the financial information search system according to one embodiment of the present invention, as a search engine, performs high-speed search by creating an index in advance by crawling a large number of search target Web pages and data at regular intervals. And using a search engine that can be a full-text search target not only for Web pages but also for documents on Web pages, and for Web pages and documents that include them for specified keywords It is a system that makes it possible to output information on the brands related to the description contents, as search results, not information on the location of.

  That is, when the search engine creates an index in advance for a group of documents such as an analyst report held on the internal system to be searched, information on the path (URL (Uniform Resource Locator)) of each document is included. In addition, an index is created by associating information on brands corresponding to the document. As a result, a full-text search is performed for the designated keyword including a document such as an analyst report, and brand information corresponding to the document referring to the keyword is output as a search result.

  This allows, for example, a salesperson to use the sales terminal to present relevant brand information to the customer on the spot, based on the keywords obtained from the conversation with the customer on the go. It is possible to use such as conducting sales activities.

<System configuration>
FIG. 1 is a diagram showing an outline of a configuration example of a financial information search system 1 according to an embodiment of the present invention. The financial information search system 1 is a search engine that uses a personal computer (PC) or a sales terminal 300 such as a tablet terminal via a network 400 such as an intranet using the Internet VPN (Virtual Private Network). 200 is a system that makes it possible to search for information including financial information held on a data source 100 held by a financial institution using 200 or the like. Here, when a full-text search is performed using keywords for documents such as analyst reports stored in the document database (DB) 120 on the data source 100 and other materials 130, as described above, the keywords The information on the brands corresponding to the documents and materials mentioned in is output as search results.

  In the present embodiment, as a search engine 200, Google (registered trademark) is assumed to satisfy the requirement that an index is created in advance by crawling and that a full-text search can be performed even in a document on a Web page. A mechanism called Google Search Appliance (GSA) provided by the company is used. By using GSA, it is possible to search for data such as in-house systems of each company using an advanced search function provided by Google (registered trademark).

  The data source 100 includes, for example, stock brand data 111 and investment trust brand data 112 which are databases that hold data such as attribute information and numerical information for each brand for each product type such as stocks and investment trusts. In addition, analyst reports, company quarterly (registered trademark), other document files such as various documents posted in association with stock codes, document DB 120 storing document data, and other related materials 130 including websites, etc. Is also included.

  The search engine built on the search engine 200 is a crawler 240 that creates an index in advance by crawling data on the data source 100 at regular intervals (for example, every 4 hours), and various indexes created by the crawler 240. Each unit includes a table and a search processing unit 250 that performs a search process on the index or the data source 100.

  The crawler 240 is provided with a plurality of crawlers 240 according to the type of data of the data source 100 in order to easily acquire or filter the search results in units of financial products or the like at the time of search. Depending on this, it should be divided into multiple parts. For example, in the example of FIG. 1, it is assumed that crawlers a to d (240a to d) are individually provided for the stock brand data 111, the investment trust brand data 112, the document DB 120, and the other materials 130 of the data source 100, As an index table created by these, the stock issue index 211 and the investment trust issue index 212 (these may be collectively referred to as “data index”), the document index 220 and the other material index 230 (collectively referred to as these). May be described as a “full-text search index”.

  In addition to the function of responding to the sales terminal 300 with a list of search results, the search processing unit 250 dynamically, for example, when a salesperson inputs a search keyword at the sales terminal 300 according to the input status. It has various support functions and auxiliary functions for the user, such as a function for suggesting keyword candidates, a function for responding to a summary of search results when a search condition is specified for the search results, and a dictionary function for keywords.

  Further, as will be described later, the financial information search system 1 uses a data source instead of a search using the index table by the search engine 200 when searching based on only tabs and other narrowing conditions without specifying a keyword. It has a DB search unit 201 for directly searching 100 data.

  In the sales terminal 300, search result information such as XML (eXtensible Markup Language) data returned from the search engine 200 is displayed as part of a search application (not shown) that displays a search screen as shown in the example of FIG. A search result processing unit 310 for processing is included.

<Create index>
Hereinafter, the contents of the index table created by the crawler 240 in the present embodiment will be described. For example, the administrator or the like can instruct the search engine 200 to set each item using a meta tag or the like as to what items are provided in the index table and how to create them. .

  FIG. 2 is a diagram showing an outline of an example of the data configuration of the stock issue index 211 as an example of the data index. The data index is created separately for each product such as stocks and investment trusts as described above in consideration of the ease of processing the search results. Accordingly, it is desirable to prepare the data of the data source 100 separately for each product and prepare the crawler 240 individually.

  The stock issue index 211 is an index table for acquiring information of a relevant issue by a search specifying a stock issue code or issue name, and the crawler a (240a) crawling the stock issue data 111 at regular intervals. And has items such as a brand code, brand name, market price, PER (Price Earnings Ratio), PBR (Price Book-value Ratio), ROE (Return On Equity), and market capitalization. The items of brand code and brand name hold the brand code and brand name set for the target brand, respectively. The items of market price, PER, PBR, ROE, and market capitalization each hold the value of each index at the time of crawling of the subject brand.

  Each of these items is set for the search engine 200 as an item to be displayed as a brand search result and an item used in a search result narrowing function (so-called faceted search). In the search result, one record of the table is a unit of one detail (brand). In addition, the basic structure of the investment trust brand index 212 and the index table of other products is almost the same, although the items differ for each product.

  FIG. 3 is a diagram showing an outline of an example of the data structure of the document index 220 that is a full-text search index. The document index 220 is created by the crawler c (240c) crawling the document DB 120 at regular intervals, and is an index table for responding a search result to a full-text search specifying a keyword. As an index used in, a document information index 220a for responding with information of a document including the keyword as a search result is included. The present embodiment further includes a document brand index 220b for responding as a search result to brand information corresponding to the document, not the information of the document including the keyword.

  FIG. 3A is a diagram showing an outline of an example of the data configuration of the document information index 220a. The document information index 220a includes items such as URL, title, analyst name, and registration date for each document. The URL item holds URL information indicating the path information of the target document. The title and analyst name items respectively hold information such as the title of the target document and the name of the created analyst. The registration date field holds information on the date on which the target document was registered / published.

  Among these items, the URL item is information for specifying the location of the target document, and the other items are items to be displayed as a search result of the document DB 120 and a narrowing function (facet search). ) Is set for the search engine 200 as an item used in ().

  FIG. 3B is a diagram showing an outline of an example of the data structure of the document brand index 220b. The document brand index 220b includes, for example, items such as URL, title, analyst name, registration date, brand code, brand name, market price, PER, PBR, and ROE for each document. Since each of these items is the same as the contents of the stock issue index 211 and the document information index 220a, the description thereof will be omitted.

  In the index, in the keyword search, when the contents of the document on the document DB 120 match, in addition to the document information of the document, information on the brand corresponding to the document is displayed as a search result, or a narrowing function (facet In order to be able to use it in (search), an item is set also about data, such as attribute information of a target brand. At this time, a brand code specifying a brand in the document information is also extracted as an index. Thus, it is possible to associate data such as brand attribute information with the brand code as a common key.

  By setting the search engine 200 to create such a document brand index 220b, when a full-text search is performed by keyword search, the index is used to refer to the specified keyword. It is possible to return information on brands corresponding to existing documents as search results.

<Screen configuration>
FIG. 4 is a diagram showing an example of a brand information search screen on the sales terminal 300. For example, the brand information search screen 320 is roughly divided into a search condition input area 321, a tab area 322, a search result display area 323, and a narrowing condition instruction area 324.

  In the search condition input area 321, the sales person can input and specify a condition for specifying or narrowing down the brand in the input field. Here, for example, when a brand to be searched is determined, a brand code and a brand name are directly designated. If the brand is not decided, a keyword (for example, “sunlight”) related to the desired theme is designated.

  At this time, for example, if you want to search directly for the detailed information of the issue of “Nomura Research Institute”, if you search by entering the character string “Nomura Research Institute”, in the normal search, In addition to the detailed information of “Research Institute”, all documents in the document DB 120 referring to “Nomura Research Institute” are also hit. Similarly, when a character string “4307” which is a stock code of “Nomura Research Institute” is input and searched, in addition to the information of “Nomura Research Institute” whose stock code is “4307”, for example, Information related to stocks with a stock price of “4307” yen and documents including “4307” in the telephone number will all be hit.

  Therefore, in the present embodiment, as shown in the search result display area 323 in the example of FIG. 4, whether the input character string is a character string indicating a brand code or brand name, or a character indicating a keyword as a search condition It is possible to identify whether it is a column or not by specifying the sales person in charge. In the example of FIG. 4, the radio button is used for designation, but the present invention is not limited to this, and an appropriate method can be used.

  When the designated character string is a stock code or a stock name, only the data system index such as the stock stock index 211 and the investment trust stock index 212 is searched when the search processing unit 250 of the search engine 200 searches the index. Thus, the search process can be performed efficiently, and only the detailed information of the brand can be output at high speed as the search result. Similarly, when the designated character string is a keyword, efficient search processing is performed only on the full-text search system index such as the document index 220, and a list of matched information is output as a search result at high speed. be able to.

  As described above, in this embodiment, the index table is created separately for each search target so that the search process can be performed efficiently. As described above, in addition to being divided according to whether the search is for detailed information (data) of the issue or full-text search, in this embodiment, the index is assigned to each product type such as stocks and investment trusts. The index 212 is created separately. Accordingly, for example, as shown in the tab area 322 in the example of FIG. 4, by providing a tab for each product type such as “stock” and “investment trust”, in accordance with the selected tab, the search target is set in advance. It is also possible to limit the index. In addition, such a proper use of the index can be realized by a “collection” function when the search engine 200 is GSA.

  In the present embodiment, as described above, even when a full-text search is performed using a keyword, information on a brand corresponding to a document referring to the keyword is output as a search result. For example, as shown in the search result display area 323 in the example of FIG. 4, the search results are displayed as a list of summary information of each brand in a state of being sorted in a predetermined order. As the sort order, in addition to the “keyword match order” shown in the drop-down list in the example of FIG. 4, various sales conditions such as “dividend yield order”, “expected PER order”, “market capitalization order”, etc. The person can give instructions as appropriate.

  As described above, for example, in a full-text search in which a keyword is specified for the document DB 120, information on brands related to the keyword is output as a search result by using the document brand index 220b including brand attribute information as a search target. can do. However, when there are a plurality of documents on the document DB 120 referring to the keyword and describing the target brand, a plurality of records of the same brand are returned as a search result. If this is displayed as it is in the search result display area 323, entries of the same brand are arranged.

  Therefore, in the present embodiment, in the search result processing unit 310 of the sales terminal 300, for the entry of the same brand included in the search result, for example, only one of the first one is adopted and the others are discarded. Perform the process summarized in

  In order to avoid sending a large amount of search results to a client terminal such as the sales terminal 300, the search engine 200 such as GSA divides the search results into a predetermined number such as every 20 items, At the time of response, only the first divided search result is responded, and thereafter, the search results are sequentially divided one by one based on an instruction from the client terminal. In this case, the search result processing unit 310 of the sales terminal 300 collectively displays the brands for the search results that are initially responded, and then responds to the next search result in response to an instruction from the sales staff. Each time, the newly collected search results are merged with the previously collected search results, and then the process of combining the issues into one is sequentially repeated.

  The search engine 200 such as GSA further has a function (facet search) for classifying search results into predetermined categories (facets) from various viewpoints and navigating the list narrowing by selecting facets. ing. The narrowing-down condition instruction area 324 in the example of FIG. 4 displays information on the hierarchical structure of facets and the number of brand data classified into each facet. By selecting a desired facet from here, the sales representative can display only the brands belonging to the facet in the search result display area 323. Note that the items used as facets are items displayed as search results set in each index table, as shown in the index table examples of FIGS. 2 and 3.

  In the present embodiment, as described above, when a search is performed by specifying a keyword for full text search in the search condition input area 321, for example, the document brand index 220b shown in FIG. By searching, a document that refers to the keyword is searched, and a brand corresponding to the hit document is displayed in the search result display area 323 as a brand related to the keyword. On the other hand, for example, in response to an instruction such as selecting the “report” tab in the tab area 322, a list of documents referring to the keyword is output as a search result as usual, instead of outputting a brand related to the keyword. It is also possible to do. In this case, the search processing unit 250 may search the document information index shown in FIG.

<Search target switching>
In the search process using the search engine 200 such as GSA, as described above, the search results of the brands are displayed after being sorted according to a condition such as “order of dividend yield”. FIG. 5 is a diagram showing an outline of an example of sorting and outputting search results by narrowing down. Here, there is a case where narrowing down is performed by specifying a tab in the tab area 322 or specifying a facet in the narrowing-down condition instruction area 324 without specifying a keyword in the search condition input area 321 of the brand information search screen 320 in FIG. Take an example.

  In general, the search engine 200 sorts and outputs in consideration of the efficiency of search processing, and a predetermined upper number of search results obtained by narrowing down (for example, 1000 in GSA) Only. In the case of a search that does not specify a keyword, only the top 1000 items based on conditions (for example, page rank (registered trademark) in GSA) are output instead of the keyword match rate, for example, a brand with a high reference frequency. It becomes a target.

  In the example of FIG. 5, among the search results, “brand AAA”, “brand BBB”, and “brand CCC” are included in the top 1000 items having the highest reference frequency, but “brand ZZZ”. Is not included, and therefore, it is excluded from the sort target in the “dividend yield order”. That is, even if the dividend yield of “brand ZZZ” is 5% and higher than other “brand AAA”, etc., the result is not included in the sort result by “dividend yield order”.

  Here, for example, when a refinement condition such as “a brand that analysts say Buy” is added, as shown in the lower diagram of FIG. 5, the previous search results are once reset and all brands are again displayed. The search is performed under the conditions narrowed down to the target, and the search result is sorted again in the “dividend yield order”.

  At this time, for example, if the number of search results that match the filtering condition is 500, which is less than 1000, the “brand ZZZ” that was not included in the top 1000 before adding the filtering condition is included. The brand with a high dividend yield such as "" can be newly included. Therefore, “sort ZZZ” is also included in the sorted search results, and when the dividend yield is higher than the others, as a result, “brand ZZZ” that has not been displayed so far is suddenly found. Can result in being displayed at the top.

  In general search, it is common sense that the search results are narrowed down sequentially by adding search conditions, but in the above example, by adding search conditions and narrowing down, There may be inconsistent results where brands that have not previously appeared as search results are suddenly displayed as having a high dividend yield. Such a situation is unlikely to be a big problem in a general search system. However, as in the financial information search system 1 of the present embodiment, the search target is a financial product, and the sales representative is based on the search result. If it is used to conduct business activities such as proposals for financial products, it may lead to violations of laws and regulations such as misleading solicitation.

  In the above example, a case where a search is performed by narrowing down without specifying a keyword is taken as an example. However, even when a search is performed by specifying a keyword, the search result is further displayed in the narrowing-down condition instruction area 324. The same situation can occur by narrowing down by designating facets. However, when a keyword is specified, when the top 1000 items are extracted, they are extracted based on the matching rate of the keyword, not the high reference frequency or page rank (registered trademark). Therefore, even if the above situation occurs, the sales representative will explain that the results are sorted with priority given to the match rate with the specified keyword among various conditions. It is possible to consider so as not to mislead solicitation.

  Therefore, in the present embodiment, when a keyword is not specified when searching for a brand, the DB search unit 201 directly searches the document DB 120 or the like without using the function of the search engine 200. . Thereby, as described above, in the refinement search of brands in the situation where no keyword is specified, the brands that can fall within the top 1000 in the sorting condition do not fall within the top 1000 in the search result before sorting. Thus, it is possible to avoid a situation caused by the restriction of the search engine 200 that is leaked from the search result after sorting.

  Specifically, when a search is instructed on the search application side of the sales terminal 300, no value is input in the input field of the search condition input area 321 of the brand information search screen 320 in FIG. In the tab area 322, narrowing is specified by a tab, facets are specified in a narrowing condition instruction area 324, and when sort conditions such as “order of dividend yield” are specified, or by keyword specification When the search is performed and the contents of the input field are cleared and the search is performed again, the DB search unit 201 is requested to perform a DB search. At this time, the search result is appropriately processed and corrected in the search result processing unit 310 of the sales terminal 300 so that the search result is displayed in a manner equivalent to the search result by the search engine 200 so that the search result is displayed seamlessly for the user. It is desirable to display it.

  In this way, not only is efficient switching performed by appropriately switching the index table used in the search engine 200 according to the specified content of the search condition, but the search of the index table by the search engine 200 and direct DB By appropriately switching between searches, it is possible to avoid a situation that may lead to a misunderstanding solicitation that a brand that should not be displayed is not displayed when sorting search results of brands that do not specify a keyword are displayed.

  As described above, according to the financial information search system 1 according to an embodiment of the present invention, when the search engine 200 creates the document index 220 in advance for the document DB 120, for example, each document In addition to the URL information, an index is created that associates information on the brands targeted by the document. As a result, it is possible to perform a full text search including the inside of each document for the specified keyword, and output information on brands corresponding to the document referring to the keyword as a search result. Based on the keywords obtained from the conversation with the customer, the sales representative uses the sales terminal 300 to present relevant brand information to the customer on the spot and conduct timely sales activities. Is possible.

  In addition, when sorting search results of brands for which no keyword is specified, the search engine 200 does not search the index table, but directly performs a DB search of the data store, so that the brand that should be displayed is not displayed. It is possible to avoid situations that can lead to solicitation.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiments. However, the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say. For example, the above-described embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to the one having all the configurations described. In addition, it is possible to add, delete, and replace other configurations for a part of the configuration of the above-described embodiment.

  INDUSTRIAL APPLICABILITY The present invention can be used for a financial information search system that searches for information related to financial products.

1 ... Financial information retrieval system,
100 ... Data source, 111 ... Stock issue data, 112 ... Investment trust issue data, 120 ... Document DB, 130 ... Other materials,
200 ... Search engine, 201 ... DB search unit, 211 ... Stock issue index, 212 ... Investment trust issue index, 220 ... Document index, 220a ... Document information index, 220b ... Document issue index, 230 ... Other material index, 240 ... Crawler, 240a to d ... crawlers a to d, 250 ... search processing unit,
300 ... Sales terminal, 310 ... Search result processing unit, 320 ... Brand information search screen, 321 ... Search condition input area, 322 ... Tab area, 323 ... Search result display area, 324 ... Refinement condition instruction area,
400 ... Network

Claims (2)

A financial information search system for searching information related to brands of financial products by a search engine via an information processing terminal via a network,
The search engine receives from the information processing terminal, a crawler that creates a full-text search index table by crawling a set of document data described for each brand held on a data source at a predetermined timing. A search processing unit that responds to a search request as a search result of a full-text search among the records related to the document data that match from the index table.
DB search that directly performs search processing on the set of document data on the data source instead of search by the search engine when no keyword is specified in the search request received from the information processing terminal A financial information retrieval system having a department. The financial information retrieval system according to claim 1,
The said information processing terminal processes and displays a search result so that it may become the display system similar to the case where it responds by the said search engine about the search result responded by the said DB search part.