JP2020086996A - Publication information retrieval system - Google Patents

Abstract

To provide a publication information retrieval system which can efficiently execute from retrieval to display for browsing of publication information.SOLUTION: A publication information retrieval system 1 comprises: a company server 2 which manages information resources collected by a user H; a company database 3 which stores the collected information resources; and a user terminal 7 which retrieves the information resources to be required by the user and displays retrieval results. The company server 2 has retrieval range limitation means for limiting a retrieval range by specifying a category relevant to the required information resources from websites A to C of other companies when the information resources to be required by the user H is retrieved, and stores the information resources matched to retrieval conditions in the company database 3.SELECTED DRAWING: Figure 1

Description

The present invention relates to a posted information search system for searching and collecting information posted on a website and listing it for use in business activities and the like.

2. Description of the Related Art Conventionally, there is known a site that picks up accommodation facility information, rental property information, job offer information, etc. that match search conditions from websites of the same industry published on the Internet. By using such a site, the user can obtain necessary information at a time without accessing the website of each company each time.

For example, the automatic recruitment information collection method of Patent Document 1 described below includes a server including a recruitment information collection unit, a recruitment company server owned by each recruiting company, and a job seeker terminal used by each job seeker. ing.

The collecting means accesses the job-related information page, receives the description information of the page, and temporarily stores it in the memory on the server. Further, the analysis unit analyzes the description information of the page acquired by the collection unit based on the specifications of a markup language such as HTML to extract a sentence and its layout information.

As a result, the latest job information that each company has published on its homepage is stored in the job information database of the server, organized by each job item component, and job seekers can immediately search for jobs from the website. Information can be browsed (Patent Document 1/Paragraphs 0020 to 0027).

JP, 2004-326712, A

However, in the automatic recruitment information collection method of Patent Document 1, since information is collected by a method called a crawler, information that is not originally necessary is also acquired, and it takes time, and the output information is used for browsing. It also took time to sort them out.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide a publication information search system capable of efficiently executing from the search of publication information to the display for browsing.

The present invention is a posted information search system that searches and collects information resources required by a user from a plurality of websites on which information resources that can be viewed by anyone are posted, and manages the collected information resources. Server, a database for storing the collected information resources, and a user terminal network-connected to the server for searching information resources required by the user and displaying search results, the server comprising: When the information resource required by the user is searched, a search range limiting means for limiting the search range by specifying a category related to the information resource from the web server of each website is provided. Is characterized by.

In the posted information search system of the present invention, a user searches for and collects necessary information resources from a website using a user terminal network-connected to a server. At this time, the search range limiting means of the server can limit the search to the category related to the information resource, and therefore, unnecessary information resources are not picked up at the time of search, and the time can be shortened. Then, the information resource that matches the search condition is stored in the database on the user side connected to the server. As a result, the user can efficiently collect the necessary information resources and browse the search results from the user terminal.

In the publication information search system of the present invention, it is preferable that the search range limiting means is capable of designating a dynamically generated page generated by an access by the user among the websites.

According to this configuration, the search range limiting means of the user terminal can specify the dynamically generated page. Therefore, the user can use the information resource posted on the dynamically generated page by searching for the information resource.

Further, in the posted information search system of the present invention, the server has an information resource inspection means for inspecting the information resource when storing the collected information resource in the database, and the information resource inspection means comprises: It is preferable to check whether or not there is duplicate information after performing the writing fluctuation correction on the collected information resources.

According to this configuration, the information resource inspecting means of the server inspects the information resource when storing the retrieved and collected information resource in the database. Since the same information resource may be posted in different forms on a plurality of websites, the information resource inspection means corrects the notation fluctuation and then determines whether or not the information is duplicate information. As a result, duplicate information resources can be efficiently found.

Further, in the posted information search system of the present invention, it is preferable that the information resource inspection unit integrates a plurality of the collected information resources determined to be duplicated information into one information resource.

According to this configuration, the information resource inspection unit of the user terminal can integrate a plurality of information resources determined to be duplicated information into one information resource. This makes it possible to eliminate duplicate information resources in the final listing.

The schematic diagram of the publication information search system concerning an embodiment of the present invention. The figure explaining the range limitation function at the time of information search. The figure explaining the data cleansing of hit information. The figure explaining information browsing of a publication information search system. The display example of the information browsing screen (web screen) of FIG. The display example of the information browsing screen (spreadsheet) of FIG.

Hereinafter, an embodiment of a posted information search system of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic diagram of a posted information search system 1 according to an embodiment of the present invention. The posted information search system 1 has a company server 2 which is a server of a company X to which an employee H who uses the system belongs, a company database 3 connected to the company server 2, and a network connection to the company server 2 through a company LAN or the like. And the user terminal 7 in which the publication information search tool (application) is installed.

Rental information, accommodation facilities, and the like can be considered as search targets, but here, job information such as regular employees and part-time jobs that anyone can view is used. That is, the company X is a company that handles job information, and the employee H (corresponding to the “user” of the present invention) uses the user terminal 7 to search and collect necessary job information from the websites of other companies in the same industry. be able to.

As shown in the figure, there are three companies, company A, company B, and company C, which have posted job postings and job advertisements for various job types on their websites. The web servers of company A, company B, and company C are referred to as management server 5A, management server 5B, and management server 5C, and their websites are referred to as website A, website B, and website C, respectively.

At this time, when the employee H of the company X inputs the search condition from the text (keyword), the check box, etc. by the posted information search tool of the user terminal 7, the company server 2 becomes the management server 5A, the management server 5B and the management server 5C. To access and search. If there is hit information (hereinafter referred to as hit information) that matches the search condition, the hit information is acquired and stored in the in-house database 3.

The in-house database 3 is a storage device such as a hard disk, and is composed of a registration data area 3A and a normalized data area 3B. In the area 3A, information such as a company name and a job type collected based on a patrol rule that patrols a website to be searched is stored. Further, in the area B, the normalized data edited from the collected information based on the normalization rule for normalizing the telephone number and the address into a predetermined format is stored.

After that, processing such as sorting is performed in the in-house server 2, and hit information is listed. In addition, since the search result list (a list screen and a spreadsheet described later) is displayed on the display screen 7a of the user terminal 7, the employee H can perform business activities based on the search result list.

Next, with reference to FIG. 2A and FIG. 2B, the features of the information retrieval by the posted information retrieval system 1, particularly the in-house server 2 will be described.

First, the range limiting function at the time of information retrieval by the in-house server 2 will be described. FIG. 2A shows the page structure (link) of the website to be searched.

The employee H inputs a keyword or an address with the posted information search tool of the user terminal 7 and searches for the job offer information. By this work, the in-house server 2 accesses the management servers 5A, 5B, and 5C of the companies A, B, and C so as to circulate. Then, the hit information matching the search condition is acquired from the websites A to C (see FIG. 1).

Here, the information search according to the present invention is different from the crawler type information acquisition in which all pages constituting a website are searched, and the category related to the information required by the employee H is specified and a predetermined range is specified. Can be limited to search.

Specifically, in the conventional crawler type information acquisition, all linked pages including the top page Ptop, the dynamically generated page Pdyn, and other pages P1 to P8 are searched.

On the other hand, the in-house server 2 of the posted information search system 1 can specify a page in a specific category as a search target. For example, the job information page of the websites A to C is a dynamically generated page that is composed of information such as a company name, a telephone number, and a URL that is determined to some extent, and when the page is switched, only the characters in the frame are replaced. .. The in-house server 2 can search only the dynamically generated page Pdyn or the page P1 in which predetermined information is posted in addition to this.

As a method therefor, there is a method of designating the type of request method, a unique attribute (region name, etc.), a speculable element (page number, etc.), and other conditions and timing of changing values. For example, the company server 2 searches the target pages 1 to 500 by specifying the attribute "tokyo", and then searches the same pages 1 to 500 by specifying the attribute "kanagawa". It is possible to perform a cyclic search process such as.

Further, the company server 2 can also be set to skip the cyclic search process under a specific condition. For example, when the information list page having a unique attribute or a speculable element in the URL is patrolled and there is no link that can be newly acquired, and the link continues for a predetermined number of pages, the patrol search process is skipped. ..

As a result, when searching for information, it is possible to avoid spending time on searching for unnecessary pages, and avoiding unnecessary capacity to acquire the capacity of the in-house database 3. As a result, it is possible to reduce the running cost of the posted information search system 1 and reduce the burden on each employee when using the system.

Next, with reference to FIG. 2B, data cleansing performed on hit information will be described.

Since the same job information may be posted on the websites A to C, job search information that seems to have been duplicated may be obtained when an information search is performed. The company server 2 (not shown) determines that the job information (e.g., company name, address, telephone number, etc.) that has been duplicated is exactly the same and stores it in the company database 3 as one piece of information.

However, there are some cases where some websites have fluctuations. In the example of the company name, the notation “Ace Co., Ltd.” is shown on the website A, but the notation “Ace Co., Ltd.” may be shown on the website B.

In addition, in the example of the address, the notation "Ochime, Chiyoda-ku, Tokyo..." is omitted on the website A, but the prefecture name "Ochime, Chiyoda-ku..." is omitted on the website B. It is possible that

In such a case, the in-house server 2 corrects the missing part of the information and the fluctuation of the notation, that is, data cleansing, before the job information is stored in the in-house database 3. After that, the job information that seems to have been duplicately hit is again matched, and if matched, it is judged as one information and stored in the in-house database 3.

Regarding telephone numbers, there are differences in the presence or absence of "- (hyphen)" and whether parentheses are used. The work of unifying the notation of telephone numbers according to such rules is called normalization and is performed as one mode of data cleansing. If this normalization is performed, the in-house server 2 can easily determine the identity of the plurality of job offer information.

Further, the in-house server 2 can specify, for each data classification, which part of the acquired page and how to extract and cleanse the data. Specifically, the in-house server 2 can individually specify the “extraction method”, “normalization method”, and “verification method” for labels (indexes) such as telephone numbers and addresses.

In the example of the telephone number, full-width numbers or half-width numbers delimited by “- (hyphen)” that appear immediately after the predetermined character string (index) are extracted, and the full-width number part is converted into half-width numbers. After that, if only the numeric part is extracted and the result is not an empty character string and there is no same telephone number in the comparison destination (telephone number already existing in the list), it is determined as a new telephone number.

In the case of a telephone number (for example, 11 digits of "09012341234"), a 3-digit number is taken out ("090", "901", "012", etc.), and the same 3 digit number as the comparison target is processed. Compare and score. Then, if they match, a high score is given, and if all the scores are equal to or less than a specified value, it is regarded as a new number.

As a result, it is possible to prevent the job information that has been duplicately hit and stored in the in-house database 3 as it is, which occurs when the websites A to C are searched. When job listing information is listed, if job listing information is duplicated, employee H may perform sales activities for the same company many times. Can be improved.

Next, referring to FIGS. 3 to 5, browsing of hit information by the posted information search system 1 will be described.

First, as shown in FIG. 3, the employee H uses the user terminal 7 (publication information search tool) to input search conditions in text or the like. FIG. 4 shows the display screen 7a of the user terminal 7, and the upper part of the display screen 7a is an example of the search designation screen.

Since "keyword" is basically used for the search in the search designation screen, the employee H inputs a keyword such as "manufacturing industry" or "accounting", and selects whether the keyword is an AND search or an OR search. Further, the date on which the job offer information is posted can be designated in the “acquisition date range”, and the work place in the job offer information can be designated in the “address search”.

When the employee H inputs search conditions from “keyword”, “acquisition date range” and “address search” and clicks the search button 7b, the own server 2 (not shown) accesses the management server 5A of the company A. , The hit information matching the search information is acquired from the website A. Although not shown here, the in-house server 2 also sequentially accesses the management server B of the company B and the management server C of the company C.

Next, the in-house server 2 stores the hit information in the in-house database 3 after executing the above-mentioned data cleansing. Further, the hit information stored in the in-house database 3 is displayed on the list screen below the display screen 7a of the user terminal 7.

On the list screen, hit information can be sorted (ascending order or descending order) by a plurality of items such as “acquisition date”, “medium name”, “company name”, “address” and displayed. As described above, since the job information that has been duplicately hit is collected as one information, the same job information is not listed on the list screen. In addition, there is a case where a company that previously issued job offer information issues another job offer information, but only the latest information is displayed.

On the list screen, the employee H can jump to the website (for example, the website A of company A) where the job listing is posted by clicking the URL of the “LINK” item. Further, the employee H can jump to the detail page in which the details of the job offer information are described by clicking the detail button 7c of the "details display" item. It should be noted that this detailed page is job information created in the company X of the employee H and stored in the in-house database 3.

In addition to the information displayed on the list screen, the job listings on the detail page include "E-mail", "Business description", "Job description", "Working hours", "Employment form", and "Treatment". ,"Salary" etc. The employee H can confirm the details of the job information of particular interest among the job information displayed on the list screen.

Further, the employee H can output the hit information stored in the in-house database 3 to a spreadsheet by clicking the export button 7d on the search designation screen of FIG. FIG. 5 is an example of a spreadsheet displayed on the display screen 7a of the user terminal 7 or printed.

In the spreadsheet, items such as “company name”, “phone number”, and “LINK” are displayed in the first row, and the hit information of this time is listed in the second and subsequent rows. The spreadsheet basically includes the same items as the list screen of FIG. 4, but the employee H may select the necessary items when outputting.

When the spreadsheet is output, the hit information can be sorted and the duplicate information is excluded, as in the list screen of FIG. Further, when the spreadsheet is displayed on the display screen 7a of the user terminal 7, it is possible to jump from the URL of the "LINK" item to the website where the job posting information is posted.

Employee H can also personalize the spreadsheet. For example, a “contacted” item for sales and a “memo entry field” item may be added. As a result, the employee H can efficiently carry out sales activities by telephone or the like to the contact information narrowed down by job type or address.

The above description is a part of the embodiment of the present invention, and various embodiments other than this are conceivable. The number of search target companies (see FIG. 1) and the website configuration (see FIG. 2A) shown in the embodiment are merely examples. The user terminal 7 is not limited to the in-house desktop PC or notebook PC, but may be a mobile device such as a tablet terminal or a smartphone that can be connected to the Internet.

Further, the information to be searched is not limited to the job offer information. For example, when the information to be searched is a restaurant, the user specifies search conditions such as “Korean cuisine” and “Setagaya Ward” from the posted information search tool of the user terminal 7. As a result, the in-house server 2 accesses the management servers of a plurality of websites that introduce restaurants (the search range is limited) and acquires hit information that matches the search information.

Further, the company server 2 cleanses the hit information and stores it in the company database 3. Then, the hit information is displayed on the list screen after performing a predetermined sort on the user terminal 7. Thereby, the user can browse the search result regarding the restaurant without accessing the plurality of websites each time and excluding the duplicate hit information.

Further, in the above embodiment, the user is allowed to access the in-house server 2, but the user may be an external company or person. In that case, the external company or person is given a specific access right to the own server 2 so that the outside company or person can search information by using the system of the own server 2. be able to. Then, the company having the in-house server 2 can develop the business of charging the usage fee and the like by providing the above service to an external company or person.

1... Posted information search system, 2... Own server, 3... Own database, 5A-5C... Management server, 7... User terminal, 7a... Display screen, 7b... Search button, 7c... Details button, 7d... Export button.

Claims (4)

A posted information search system that searches and collects information resources required by users from a plurality of websites on which information resources that can be viewed by anyone are posted,
A server for managing the collected information resources,
A database storing the collected information resources,
A user terminal that is network-connected to the server and that searches for information resources required by the user and displays the search results;
The server has search range limiting means for limiting a search range by specifying a category related to the information resource from the web server of each website when searching the information resource required by the user. A posted information search system that is characterized by: The publication information search system according to claim 1, wherein the search range limiting means is capable of designating a dynamically generated page generated by an access by the user among the websites. The server has an information resource inspection means for inspecting the information resource when storing the collected information resource in the database,
3. The posted information search according to claim 1 or 2, wherein the information resource inspection means inspects whether or not there is duplicated information after correcting notation fluctuations of the collected information resources. system. The posted information search system according to claim 3, wherein the information resource inspection unit integrates a plurality of the collected information resources determined to be duplicated information into one information resource.