JP2000215206A - Homepage search system - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To provide a homepage search system that allows a user to more accurately obtain desired information by performing a stepwise search between a center side and a terminal side. A homepage search system that searches for a desired homepage on the side of a center by transmitting search data from the terminal to a center, and transmits the search result to the terminal. The terminal 100 searches for a homepage address and a text portion in the homepage associated with the primary search data sent from the server, and transmits the result to the terminal 100 as a primary search result. The second search means 120 sends the primary search data to the center 200, receives the primary search result for the primary search data, narrows the result using the secondary search data, and extracts the secondary search result. , To be realized.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technology for searching a homepage on the Internet.

[0002]

2. Description of the Related Art In recent years, various homepages (information) have been disclosed on the Internet. For example, by accessing the Internet using a personal computer, it is impossible to obtain desired information. is there. For this reason, the number of users who use the Internet is increasing even when obtaining necessary information at home.

[0003] When information is obtained from the Internet in this manner, a site (Web site) where the information that the user desires exists exists.
b site / home page address / URL (Unifo
rmResource Locator)) often lacks clues, and in such cases, use search engines such as "Yahoo" (Yahoo Japan Corporation) and "Lycos" (Carnegie Mellon University) published on the Internet. By doing so, desired information candidates (search results by specifying keywords) are obtained in the URL list of the homepage, and the desired information (homepage) is obtained by referring to (accessing) those URLs.

However, in the conventional homepage search using a search engine published on the Internet as described above, the communication line needs to be connected even while the search engine is performing a search process (accurately, the Internet). Connection at home), and there is a financial problem due to communication charges in home use using a general public line that is charged on an hourly basis.

[0005] Therefore, a technique for solving the above problem is disclosed, and a "homepage search system" disclosed in Japanese Patent Application Laid-Open No. 10-154152 is one example.

In the "homepage search system" disclosed in Japanese Patent Application Laid-Open No. H10-154152, communication processing of a search code (search keyword or the like) transmitted from a terminal (user-side device) to the center is performed. The communication processing of the search result transmitted from the center to the terminal is made independent on the communication line. That is, it is characterized in that the terminal has a configuration in which no problem occurs even when the terminal is not connected on the communication line from when the search code is transmitted until the search result is obtained.

[0007]

However, the technique described above still has the following problems.

[0008] In the "homepage search system" of Japanese Patent Application Laid-Open No. H10-154152, the data format of the search result transmitted from the center is unclear, and from the disclosed contents, the data format of the search result is (1) ) URL list (the title of the homepage is assumed to be added) (2) It is presumed that one or both of the lists of the homepage data itself are included.

First, when the data format of the search result is (1), there is a problem that "the search result cannot be further narrowed down only by the terminal (without connecting to the center again)". .

[0010] As described above, it is no exaggeration to say that various types of information are disclosed on the Internet, and that desired information always exists somewhere. Considering this conversely, it also means that there is a very large amount of information that is not required.

That is, in a homepage search,
The number of homepages that hit in one search becomes enormous (conversely, if the number of hits is small, the search engine itself may be a problem), and the URL obtained from the search results and the title information of the homepage alone will be a list. Refers to (accesses) the homepage of the first candidate URL that was found, checks the contents of the homepage, and as a result,
If the information is not what you want, the same process is performed for the next candidate until the target candidate is found, so no matter how much call time is reduced during the search process,
It takes time to access the homepage and check the content of the homepage information, and after all, the communication (internet connection) time may not be reduced.

Next, when the data format of the search result is (2), "(2-1) there is a possibility that the data amount of the homepage itself is large, (2-2) a copyright problem".
There are two points.

[0013] Regarding the above (2-1), not only text data but also information of various data formats such as images and sounds exist on one homepage. In particular, data such as images and sounds has a large data amount. Naturally, downloading the data requires a long communication (Internet connection) time. Regarding the item (2-2), there are a lot of information in the homepage that the publisher has copyright. In particular, image, video, and audio information has almost all copyrights. That is, providing the home page data to the user at least as it is via the center is considered to be a copyright problem.

The present invention has been made in order to solve the above-mentioned problems. It is an object of the present invention to carry out a first-stage search for extracting only a text portion of homepage data at a center side. By executing the second search, which further narrows down the search result on the terminal side, the user can obtain desired information accurately. At that time, the amount of information to be searched can be reduced. It is an object of the present invention to provide a homepage search system which can reduce the search time and can be realized without the terminal and the center having to constantly communicate and connect between the request instruction and the result acquisition.

[0015]

According to a first aspect of the present invention, a terminal is connected to a center by a communication line, and a terminal is connected to the center by a terminal.
A homepage search system that searches for a desired homepage on the center side by transmitting search data and transmits the search result to the terminal, wherein the center is a homepage search center, and Communication means for transmitting and receiving data and connecting to the Internet, storage means for storing search data and search results transmitted from the terminal, and transmitting the search data from the terminal to the communication means. , And extracts a homepage address and a text portion in the homepage associated with the search data as a search result, and stores the search result in the storage unit.
First search means for transmitting a first search result to the terminal, wherein the terminal is a homepage search terminal, and communication means for transmitting and receiving data to and from the center; Input means for inputting data required by a user for search, display means for displaying data input by the input means or a search result, data input by the input means and search A storage unit for storing a result, and data for a primary search at the center input by the input unit are transmitted to the center via the communication unit, and the search result is transmitted to the first center via the communication unit. Control means for receiving the result as a next search result and storing the result in the storage means; and a second search result from the center stored in the storage means, the second search result being inputted by the input means. The narrowing by using the data for the search 2
And a second search means for extracting a next search result, wherein a step-by-step search of the homepage is performed.

According to a second aspect of the present invention, in the homepage search system according to the first aspect, the first search means searches for a homepage address related to search data from the terminal. Search means, search result processing means for extracting a text part in the homepage of the homepage address, word dictionary, first keyword extraction means for converting the text part into a plurality of keyword data using the word dictionary, And outputting the keyword data, which is the conversion result of the first keyword extraction means, in the first search result.

According to a third aspect of the present invention, in the homepage search system according to the first aspect, the second detecting means is inputted from the input means using the word dictionary and the word dictionary. Second keyword extracting means for converting the data for secondary search into a plurality of keyword data, wherein the keyword data of the primary search result by the first keyword extracting means and the output by the second keyword extracting means It is characterized in that the results are converted into word vectors, comparison processing is performed, and the degree of coincidence is determined.

According to a fourth aspect of the present invention, in the homepage search system according to the second aspect, the primary search result output from the first search means is at least a URL which is a homepage address.
It is characterized by comprising a character string, a title of the page, a text summary of the page, and a keyword included in the page text summary.

A homepage search system according to a fifth aspect of the present invention is the homepage search system according to the first aspect, wherein the first search means searches for a homepage address related to search data from the terminal. It has a search means and a search result processing means for extracting a text portion in the home page of the home page address, and outputs the extracted text as a primary search result.

According to a sixth aspect of the present invention, in the homepage search system according to the first aspect, the second detecting means includes a word dictionary and a first order from a center stored in the storage means. First keyword extracting means for converting a text portion of the search result into a plurality of keyword data using the word dictionary, and data for a second search input from the input means using the word dictionary by a plurality of keyword And a second keyword extracting means for converting the keyword data which is an output result of the first keyword extracting means and a keyword data which is an output result of the second keyword extracting means into word vectors. Then, the degree of coincidence is determined.

According to a seventh aspect of the present invention, in the homepage search system according to the fifth aspect, the first search result output from the first search means is at least a homepage address URL.
It comprises a character string, a title of the page, and a text summary of the page.

According to an eighth aspect of the present invention, in the homepage search system according to the first or second or fifth aspect, the first search means is configured to execute the first search when the first search is completed. It is characterized in that the mail address received from the terminal is notified that the primary search at the center has been completed.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a homepage search system according to the present invention will be described below with reference to the drawings.

(First Embodiment) FIG. 1 is a functional block diagram showing a system configuration of a homepage search system according to a first embodiment of the present invention.

This system comprises a terminal 100 operated by a user to obtain desired information, and a terminal 100
And a center 200 for transmitting information such as a site (homepage address) where it is thought that related information exists based on a keyword requested by a user through the Internet.

The terminal 100 displays an input device 102 for inputting a search keyword or the like or instructing an operation, and data and processing input from the input device 102, here information such as a search result. Display device 101 for
And communication for controlling line connection between the storage device 103 as a work memory for storing input data and search results and for control processing required for the terminal 100 to operate, and the center 200. A control unit 110 for controlling each of the device 104 and the display device 101, the input device 102, the communication device 104, and the storage device 103;
The control means 110 transmits the search keyword (first key a1) input from the input device 102 to the center 200 as mail information (mail address first key a2) by controlling the communication device 104. Receives the result of the processing in the center 200 (first search result a3) and stores the information (first search result a4) in the storage device 103.
Control the memorized in.

Further, target information is retrieved / extracted and controlled with respect to the processing result (first retrieval result a5) at the center 200 received from the center 200 which is a feature of the terminal 100 and stored in the storage device 103. Search means 1 for
20. This second search means 120
Is a word dictionary 125 in which various keywords for use in the terminal 100 are stored, and a keyword (second key a6) input from the input device 102 is stored in the word dictionary 12
Second keyword extraction means 122 for searching within
And the result of the processing at the center 200 stored in the storage device 103 (the first search result a
The search means 121 performs a process of searching and extracting target information for 5) and returning a result (second search result a8) to the storage device 103.

Here, the input device 102 may be a keyboard, a pointing device such as a mouse, or a touch panel including a tablet. The display device 101 may be a CRT, a liquid crystal display device such as an LCD, or a liquid crystal tablet integrated with the input device 101, for example, a display integrated type. The communication device 104 may be a device having a function of connecting to a general public line, for example, a device capable of connecting to the Internet. The storage device 103 may be a semiconductor memory element (for example, a nonvolatile memory such as a flash ROM or an EEPROM, or a RAM if backed up by a power supply), a hard disk or an optical disk (a magneto-optical disk, an MD, a DVD, or the like). And the like. The actual hardware configuration of the control unit 110 and the second search unit 120 is a CPU, a RAM, a ROM, and the like.
And the like, and these may be dedicated LSIs or may be configured in combination.

Next, the center 200 has a communication device 201 for connection with the terminal (service user) 100 and for connection with the Internet (mail server) 300, and a search transmitted from the terminal (service user). A storage device 202 as a work memory for saving keywords and search results and for performing control processing required for the operation of the center 200; In addition to controlling the device 202, a related site (homepage address) is searched based on a search keyword transmitted from the terminal (service user) 100, and text data in the homepage of the related site that has been searched, or Extracts word keywords of text data and hits homepage address information Addition to the first detection means for sending requested terminal (service user) to 100 210
Is provided.

Here, the center 200 receives the search keyword transmitted from the terminal (service user) 100 via the communication device 201, and receives the search keyword (first key b1). The first search means 210 searches for a related site (homepage address) according to the search result, and extracts text data or a word keyword of the text data in the homepage of the related site that has been searched for. The information (first search result b2) is transmitted to the requesting terminal (service user) 100 in addition to the homepage address information, and the information (first search result b3) is also stored in the storage device 202. I do.

The first search means 120 is provided by the center 20
0, based on a word dictionary 216 storing various keywords to be used in the terminal 100 and a keyword transmitted from the requested terminal 100, a communication destination specific address (Web site address / Web site address /
URL: Uniform Resource Local
related URL search means 211 for searching for tor), a URL database 215 to which the related URL search means 211 performs URL search, and a related URL search means 21
1. A search result processing unit 212 that processes information (URL list b4) as a search result according to No. 1 and processes (generates) text data b5 included therein, and a search result processing unit using the word dictionary 216. And a first keyword extracting means 213 for extracting a word keyword from the text generated in 212.
3 is data obtained by adding the hit homepage address information to the word keywords of the extracted text data,
Alternatively, data obtained by adding the hit homepage address information of the text data to the search result information (first
It has a function of storing the search results b2, b3) in the storage device 202, giving the search results b2, b3) to the communication device 201, and transmitting them to the requesting terminal 100.

Here, the communication device 201 may be a device having a function of connecting to a general public line, for example, a device capable of connecting to the Internet.
The storage device 202 may be a semiconductor memory element (for example, a nonvolatile memory such as a flash ROM or an EEPROM, or a RAM if backed up by a power supply),
Hard disk and optical disk (magneto-optical disk, MD,
An auxiliary storage device such as a DVD may be used. The actual hardware configuration of the first search unit 210 is assumed to be configured by a CPU, a RAM, a ROM, and the like, and may be a dedicated LSI or a combination thereof. Good.

The word dictionary 125 in the terminal 100 and the word dictionary 216 in the center 200 have the same contents.

In this embodiment, the related URL search means 211
The search result processing means 212 and the search result processing means 212 are configured separately, but this is only to make the process and function of the processing in the center 200 easy to understand. In addition, the [conventional technology]
In comparison with Japanese Unexamined Patent Publication No. Hei 10-154152, "Homepage search system", the related URL search means 211 corresponds to "search means". Are performed in parallel (within the same means).

As described above, the terminal (service user) 100
It may be a desktop or handheld (including a notebook) personal computer, a portable information tool (including a portable information terminal, a communication terminal, etc.) that has recently become widespread, or The terminal device may be provided with a workstation function including the above. In addition, the center 200 may be a device provided with a storage unit capable of constructing a database having a relatively large capacity, for example, a dedicated information processing device or a general-purpose workstation.

The center 20 described with reference to FIG.
0 and the respective units constituting the respective operations of the terminal 100 will be sequentially described.

(1) Operation on the Center Side FIG. 2 is a flowchart showing the operation of the related URL search means 211.

The operation of the related URL search means 211 is as follows.
As described above, it is basically the same as the "search means" of Japanese Patent Application Laid-Open No. H10-154152 "Homepage Search System" described in the section of [Prior Art].

First, when the center 200 is started, the related URL search means 211 keeps watching through the communication device 201 whether or not a search request has been issued (S201). Is detected, the process proceeds to the next step (S202). Since a search request has occurred, “user name (ID)”, “password”,
Information such as “mail address” and “mail transmission presence / absence information” is received (S203). A folder unique to the user for storing the received information is generated (that is, a storage area required for one search request from the terminal 100 is secured in the storage device 202) (S204). The received information such as "user name (ID)", "password", "mail address", and "mail transmission presence / absence information" is stored in the secured storage area in association with each other (S20).
5). Further, it receives “search keyword” and “upper limit of additional information capacity (upper limit of capacity size of area to be temporarily reserved)” (S206). A search is made in the URL database 215 using the received search keyword as a designated keyword (S207). A URL list, which is the relevant information obtained by the above search, is output as a search result. The output URL list of the search result is provided as a parameter to the search result processing unit 212 (S208).

The above is the operation of the related URL search means 211. The steps S202 to S205 are mainly for managing search results for each user, that is, for each search request from the terminal 100. This is because service charge management and search results are managed by another person other than the search requester. This is because it is necessary to protect from.

FIG. 3 is a display example of page data when the search service is disclosed on the Internet.

This page data is stored in the terminal 1 when the user operates the terminal 100 to access a desired site.
00 is displayed on the display device 101. It should be noted that the description of “for center” in the title character string
Is used to access the center 200, which means “primary search” at the center 200.

On the screen, various operation buttons and various data (user name (ID), password, mail address, mail transmission information, search keyword, etc.) described in FIG. 2 are input. A data entry field is provided.

An acquisition button 401 is a button for downloading a search result at the center 200, and a start button 402 is a button for starting a search at the center 200. When one of the buttons is pressed, the "user "Name" (404), "Password" (405), "E-mail address" (406), "E-mail transmission presence / absence information" (Information indicating whether direct result transmission at the time) (407), "keyword"
The content of each field of (408) is a page transmitted to the center. If the user wants to end the process halfway, the user can press the end button 403 to execute the process.

Here, when the search result in the center 200 is transmitted by e-mail, not only the “keyword” (405) field but also the “user name” (40)
If the start button 402 is pressed without inputting the fields 4) and "Password" (405), the search service cannot be received (the center 100 notifies the terminal 100 that an error has occurred). Will be. Also, when the search result in the center 200 is directly transmitted at the time of the request by the acquisition button 401, the fact that the field input of the "user name" (404) and the "password" (405) is required is the same. It is. In other words, the processing of FIG. 2 is executed by transmitting these pieces of information from the terminal 100 to the center 200.

The input of various data and the operation of buttons are performed by using the input device 102.

The screen shown in FIG. 3 is an example of an implementation method for user management, and is not limited to this. For example, another screen, for example, the terminal 100 shown in FIG. 1 is shown. However, if a device that can access an ID card (a card such as a magnetic card, an IC card, or an optical card) is provided, the user can insert the ID card into a read access device (not shown) to perform “usage”. Person name "
(404), “password” (405), and “mail address” (406) may be automatically set.

FIG. 4 shows the related URL search means 211 of FIG.
FIG. 4 is an explanatory diagram showing the structure and contents of a URL list as output contents (search results) of the URL.

FIG. 4A shows a structure (structure) of a URL list, and FIG. 4B shows an example of data actually stored in the structure (structure).

Here, the number of items stored in the URLList is the number of items for the variable MAXCNT.
As the structure of LList, a character string of the communication destination unique address of the home page as the URL (MAX 256 characters)
And a character string (MAX256 characters) of the title name of the page as Title.

The actual data stored in the structure of this URL list is URL = “h” in the case of URLList [0].
http: // www. space. xxxx ", Ti
tle = “About Japanese space development”. In addition, NULL data is stored in an array of element Nos in which no data is stored. In this example, the maximum length of the character string is 256, but the present invention is not limited to this.
2 depends on the capacity.

The related URL search means 21 of FIG.
The URL list, which is the output content (search result) of No. 1, is given to the search result processing means 212 as a parameter.

FIG. 5 shows the search result processing means 212 and the first
6 is a flowchart illustrating an operation of a keyword extracting unit 213.

These units process the output contents (assumed to be delivered in the work area URLList []) of the related URL search unit 211 in FIG. 4 into a data (FstResult) format to be delivered to the user, and
2 is stored in a storage area provided for each user, and this data is transmitted in response to a request from the terminal (service user) 100 or by e-mail.

FIG. 6 shows a first search result (FstResul) delivered to the user as a processing result of the first search means 210.
It is explanatory drawing showing the structure and content of t []).

First, the contents will be described with reference to FIG. FIG. 6A shows a structure (structure) of the first search result, and FIG. 6B shows an example of data actually stored in the structure (structure).

Here, the number of cases stored in the FstResult is the same as the number stored in the URLList in FIG.
As the structure of tResult, a character string of the communication destination unique address of the home page (MAX256 characters) as the URL, a character string of the title name of the page (MAX256 characters) as the Title, and an outline (summary) of the text of the page as Abst (MAX 256 characters) and a keyword character string (MAX 1024 characters) included in the summary (summary) of the text of the page. However, URL and Title are the same as those in FIG.

When the actual data stored in the structure of the first search result is FstResult [0], the URL =
In “http: //www.space.xxxx”,
Title = "About Japanese Space Development", Abst
= "Space development is underway by the Japan Space Development Agency (the central executing agency of Japan's space development) ...", Key
word = “Japan, space, development, business, institution, rocket, ...”. In addition, NULL data is stored in an array of element Nos in which no data is stored. In this example, the maximum length of the character string is 256 and 1
024, but is not limited to this, and its size depends on the capacity of the storage device 202 of the center 200.

In the above-described embodiment, the data is stored in a storage area provided for each user in the storage device 202 in FIG. 6, and the data is stored in response to a request from the terminal (service user) 100 or by mail. Sent.

Next, the related URL search means 21 shown in FIG.
FIG. 5 shows how the output contents (URL list: URLList []) of No. 1 are processed into the first search result (FstResult []) passed to the user in FIG.
This will be described with reference to the flowchart of FIG.

The URL list (URLList []) shown in FIG. 4 as an input parameter to the search result processing means 212.
Is given, local variables used for processing are initialized (i = 0) (S221). URLLis
The search is sequentially performed with reference to the contents of the member variable URL in t []. The variable URLList [i]. It is checked whether or not the address of the homepage is stored in the URL (S222). If it is stored, the process proceeds to the next step S223. If it is not stored, the process proceeds to step S2.
Proceed to 28. The variable URLList [i]. When the homepage address is stored in the URL, the communication device 2
01 via the variable URLList [i]. The user accesses the home page indicated by the URL and downloads the contents (S223). This access and download are performed via the Internet mail access b12 in FIG. The data of the downloaded home page is HTML format data.

For example, FIG. 7 shows an example of downloaded homepage data.

FIG. 7 shows the URL of element No. 1 in FIG.
URLList [1]. URL = “http
p: // www. shuttle. http: //www.xxxx ". This is the homepage data downloaded when accessing" xxx. "
>, <Title>, </ title>, <body
>, </ Body>, </ html> are stored.

Next, a text portion is extracted from the downloaded HTML home page data (S2).
24). There are various methods for extracting the text part of the homepage data, including extracting the linked homepage and extracting the text part around the part that matches the search keyword for the limited characters. However, the simplest method is to extract the contents of the <TITLE> tag and the text part (content part) of the <BODY> tag.

Then, the first keyword extracting means 213 converts (extracts) the extracted text data into a plurality of word keywords using the word dictionary 216 (S).
225). In this embodiment, this step S225 is referred to as a first keyword extraction step, and the keyword group extracted by this means is described later (processed on the terminal 100 side).
Used for word vector generation. For example, if the extracted keyword group is element No = 1 in FIG.
keyword = “space shuttle, airplane, Japan, mission,...”.

The structure of the first keyword extracting means 213 and the contents of its processing are described in Japanese Patent Application Laid-Open No.
The technology disclosed in US Pat. No. 166611, “Data Searching Apparatus Based on Natural Language” is used. Therefore, description of specific contents is omitted (this step S
225 corresponds to the first half of S22 in FIG. 2 of Japanese Patent Application Laid-Open No. 8-16611, entitled "Data Retrieval Device Based on Natural Language".

Including the result of step S 225,
FstResul
Stored in t [i]. The data of the FstResult [i] is stored in a user-specific storage area allocated to the storage device 202 in advance (the first search result b in FIG. 1).
3). The user-specific storage area is a user-specific folder that is generated and secured in advance in step S204 in FIG.

That is, the variable URL where the hit home page address which is the output data (first search result) of the first search means 210 described in FIG. 6 is stored, <TI
TLE> Variable Title in which contents of tag are stored, <B
ODY> The beginning text content of the tag (in this example, the beginning 25
Abst, <TITLE> storing 6 bytes)
A FstResult [] composed of four members of a variable Keyword in which the result of extracting the contents of the tag and the <BODY> tag by a word keyword is stored.

In the processing from step S222 to S226, the local variable i is incremented (S22
7) Then, the process is repeated and executed sequentially for the number of items registered in the URL list of FIG.

When the processing for the number of items registered in the URL list of FIG. 4 is completed, the processing is performed on the terminal (service user) 100 that has requested the first search result, which is the processing result of the first search means 210. A transmission method is determined, and a process of returning a result is performed by the method.

In step S203 of FIG.
In step S205, the data of the mail address 406 stored in the user-specific folder of the storage device 202 is confirmed (S228). As a result, when the mail address does not exist, the processing in the center 200 is temporarily terminated. On the other hand, if the mail address exists, the completion of the first search processing in the center 200 is notified to the mail address (S229), and further received in step S203 in FIG.
The mail transmission presence / absence information 407 stored in the user unique folder 02 is confirmed (S230). As a result, if the terminal (service user) 100 has not requested transmission of the first search result by e-mail, the processing in the center 200 is temporarily terminated. On the other hand, the terminal (service user) 100
Requests transmission of the first search result by e-mail, FstResult which is the first search result in FIG.
The contents of [] are transmitted by e-mail (S231). These correspond to the path of the first search result b2 in FIG. When the first search result is transmitted by e-mail, in the present processing flowchart, it appears that the completion notification and the data transmission are individually performed. However, as the completion notification mail with the first search result data file attached, May be processed. After the transmission of the mail is completed, the process proceeds to step S236.

On the other hand, as a result of the determination in S228 or S230, if the first search result is not transmitted by mail, the system keeps watching whether or not a request for obtaining the search result has been issued (S232). If it is detected that the acquisition request has occurred, the process moves to the next step (S233). This search result acquisition request is sent to the terminal (service user) side 100 during the search processing at the center 200.
This function is used to enable the disconnection of the line connection with 0, and is used only by the search requester to obtain the search result at the center 200. The center 200 receives the "user name (ID)" and the "password" when the retrieval result acquisition request is issued (S233). In the center 200, the transmitted "user name (ID)" and "password"
Terminal from which search results should be obtained (service user) 10
It is checked whether it is 0 (S234). As a result, if it is determined that the terminal is not the correct terminal (service user) 100, the termination processing is performed without any operation. On the other hand, when it is determined that the terminal is the correct terminal (service user) 100, the corresponding FstRe, which is the first search result of FIG.
The content of "sult []" is transmitted (S235).

In this embodiment, when the first search result is transmitted to the requesting terminal 100, it is determined that the same search result will not be used again thereafter, and the search result (FstResu
lt []) and usage management information are deleted from the storage device 202 (S236). By this, Center 2
00, the storage area of the storage device 202 can be effectively and effectively used. If the storage device 202 has room, a processing method that does not delete the data may be used.

(2) Operation on Terminal (Service User) Side FIG. 8 is a flowchart showing the operation of the control means 110.

First, the terminal 100 is activated, and a process of connecting to the center 200 is performed via the communication device 104 (S101). When the line is connected, an input screen (for example, the screen in FIG. 3) for inputting various parameters for a search request to the center 200 is displayed on the display device 101 (S102). When the search service of the center 200 is a service published on the Internet, screen data (HTML data format) as shown in FIG. 3 is displayed by Internet access to a site (URL) desired by the user. Download and view.

Thereafter, various parameters for each search are input by the user using the input device 102 on the displayed input screen, and the fields (404 to 408) shown in FIG. ), And the settings are completed,
The input contents are stored in each field variable and display processing is performed (S103, S104).

At this time, it is also determined whether or not a button for starting the process has been pressed. First, Center 200
It is determined whether or not the start button 402 for requesting the user to start the search is pressed (S105). If it has been pressed, the process proceeds to step S106, and if it has not been pressed, it is determined whether or not an acquisition button 401 for acquiring a search result has been pressed (S111). If it has been pressed, the process proceeds to step S112. If it has not been pressed, the process waits until the next input operation is performed (that is, step S1).
Return to 03).

If the start button 402 for requesting the center 200 to start a search is pressed, a search request command is transmitted to the center 200 (S1).
06), “user name (ID)” (404), “password” (405), input in step S104,
The parameters for the search of “mail address” (406) and “mail transmission presence / absence information” (407) are stored in the center 20.
0 (S107), and further transmits a "keyword" (408) for searching at the center 200 (corresponding to the first key a2 in FIG. 1) (S108). At the time of transmission, each of the above data may be transmitted separately or, for example, may be transmitted in a form attached together by e-mail.

After the transmission, it is determined whether the method of obtaining the first search result from the center 200 has been set to be acquired by mail (S109, S110).
The process proceeds to step S115 where the end processing (line connection, end processing) is performed unconditionally.

In this case, the contents are checked as soon as the mail arrives. If the contents are the desired contents, the first search results (first search results a3 and a4 in FIG. 1) from the center 200 are inputted at the time of the request. The keyword is stored in the storage device 103 in association with “(408)”.

When the search result is directly received from the center 200 by setting not to be obtained by e-mail, in the present processing flowchart, a route is provided to shift to the input step of each parameter (represented by a dotted line). )
However, the processing step of the search start request and the processing step of the search result request described later are configured independently, and the end button 4
For example, a route (path) for terminating the process halfway (line connection, termination process) may be provided, such as 03.

On the other hand, if the acquisition button 401 for requesting the search result (first search result) to the center 200 has been pressed, a search result acquisition request command is transmitted to the center 200 (S112). Center 20
In order to confirm whether the user who made the request at 0 and the result requester are the same, the “user name (ID)” (40
4) Transmit the contents of "password" (405) (S
113). As a result, first search results transmitted from the center 200 (first search results a3 and a4 in FIG. 1).
Is stored in the storage device 103 in association with the “keyword” (408) input at the time of the request (S114). still,
In this embodiment, the reception date of the first search result is also stored in the storage device 10.
3 is stored. When the process is complete,
Finally, the connection with the center 200 is disconnected (S11).
5).

In this embodiment, the time when the terminal 100 starts the line connection to the center 200 is the time of the above-described step S101. However, the time when the line connection is turned on is not limited to this. , Step S10
5, when the start button 402 is pressed,
It may be the time when the acquisition button 401 is pressed in 11.

FIG. 9 is a flowchart showing the operation of the second search means 120.

The second search means 120 is started after the control means 110 disconnects the line connection with the center 200 shown in FIG.

First, an input screen for inputting various parameters for the secondary search is displayed on the display device 101 (S121).

FIG. 10 is a display example of the secondary search screen display.

FIG. 10A is an input screen for inputting various parameters for the secondary search, and FIG. 10B is a screen displaying a search result based on the input parameters.

The input screen for inputting various parameters for the secondary search shown in FIG.
01, a search key input field 502, and a check box 503 for selecting whether or not to execute a secondary search for each first search result. In the search key input field 502, the user inputs a desired keyword using the input device 102. The search key input field 502 has a specification that allows input in a natural language as shown in FIG. In the display area of the “center search list”, the date and time when the result has been obtained from the center 200 and the first search result stored in the storage device 103 has been stored, and the keyword at the time of requesting the center 200 to perform the search. Is displayed, and a check box 503 for selecting whether or not to execute the secondary search for each first search result is provided at the tip thereof.

In the flowchart of the second search process shown in FIG.
Is selected (S125), the presence / absence display of the check box 503 to be searched is inverted according to the operation (S1).
26) Return to the input waiting state.

On the other hand, when it is determined that a search key has been entered in the search field 502 (S122), the input data is stored in a variable for this field, and the input content is displayed (S123).

Further, the contents entered in the search field 502 are converted (extracted) into a plurality of word keywords by the second keyword extracting means 122 using the word dictionary 125 (S124). In this embodiment, this step S124 is referred to as a second keyword extraction step, and the keyword group extracted by this means is used for generating a word vector described later.

This process is executed immediately after the search key is input to the search field 502 (S123).
This only corresponds to FIG. 1 (a block diagram showing the configuration of the homepage search system), and may actually be a step that is performed immediately after it is determined that the start button 501 has been pressed (S127) ( Just because you enter a search key doesn't necessarily mean that you start searching.) However, in the case of FIG.
Immediately after it is determined that 01 has been pressed (S127), batch conversion is performed, and any method may be used.

The structure of the second keyword extracting means 122 and the contents of the processing steps are the same as those described in the description of the first keyword extracting means 213, as disclosed in Japanese Patent Application Laid-Open No. Hei 8-16611. The technology disclosed in the official gazette “Data retrieval device based on natural language” is used. Therefore, the description of the specific contents will be omitted (this step S124 is described in S22 of FIG.
Of the first half).

Next, when the start button 501 is pressed (S127), it is confirmed whether or not a search target is set in the check box 503 (S128). If “Yes” is not set for any of the check boxes 503, the process returns to the state of waiting for input on the original input screen. If “Yes” is set for any of the check boxes 503, the actual secondary search process is executed for all the set search target data.
That is, the repetition processing is performed until the primary search results listed in the display area of the “center search list” in FIG. 10A with the check box 503 turned on sequentially disappear (S128). ), And the result is the screen display shown in FIG. 10B after the completion of the secondary search processing.

The processing from steps S129 to S135 is a result of the primary search for one of the search target data, that is, the result of the primary search in the center 200.
FstResul to be searched stored in 03
This is a sequence of processing for performing a secondary search for t [].

First, local variables for secondary search are initialized (i = 0) (S129). Fst of the above search target
Result [], URL list data FstResult [i]. The URL is checked (S130). FstResult [i]. URL
If there is data in the keyword group, the keyword group FstResu
lt [i]. Each word data of the keyword is converted into a word vector (S131). For example, FstResult [0]. Each of the word data such as "Japan", "space", "development", etc. in the Keyword is vector-quantized. Next, each word data similarly configured for the search key input by the user in steps S122 to S123 for the secondary search is converted into a word vector (S132). For example, regarding the search key “I want to know information about Japanese astronauts” in FIG.
Vector quantization of each word data such as "Japanese", "space", "astronaut", etc. is performed. The degree of coincidence between the vectors calculated in steps S131 and S132 is arithmetically (compared) (S133). Judge that there is, FstResult [i]
, That is, the information of the variables URL, Title, and Abst are sequentially displayed (S134). After that, the local variable i is incremented (S135), whereby Fst
Result [i]. The processing is sequentially repeated and executed for the number of cases registered in the URL list of the URL.

[0098] In step S130, FstRes
ult [i]. If there is no more data registered in the URL list of the URL, it means that the processing for the primary search result of the search target has been completed, and the process returns to the upper step S128, where the primary search target which is still the search target is returned. Check if there are any more search results. at the time,
If it still exists, the processing of step S129 and subsequent steps are performed in the same manner, and if it does not exist, the secondary search result is displayed in FIG.
The screen is displayed as shown in FIG.

Also, in the processing flow of FIG.
It is described that the contents of the result “sult [i]” are sequentially displayed in step S134.
At the time of step S134, a mark indicating whether or not the data is data to be displayed is attached (for example, a flag area is provided and a mark is set there), and actually, "NO" is determined in step S128. At this point, a method of displaying all the marked data at once may be used.

When the user wants to end the process, the user clicks the “end” button in FIG.
It is executed by pressing the button.

The vectorization (vector quantization) processing of word data in steps S131 and S132 is disclosed in Japanese Patent Application Laid-Open No. 8-16611, entitled "Data Retrieval Apparatus Based on Natural Language", which has been filed earlier. We will use the technology that we have. Therefore, the description of the specific contents is omitted. (This processing is described in Japanese Patent Application Laid-Open No. 8-16611, entitled "Data Searching Apparatus Based on Natural Language" in FIG.
In the second half and S23).

As described above, by executing the configuration, function, and processing of the present embodiment, the conventional homepage search system could not obtain: In addition to further shortening the communication time (Internet connection time), Searching (narrowing down) is possible.-Since the information on the homepage is not provided to the user as it is, copyright issues are reduced.-Searching in natural language (users enter keywords in their own words). Can be performed.

(Second Embodiment) FIG. 11 shows a second embodiment of the present invention.
It is a functional block diagram showing the system configuration of the homepage search system in the embodiment of FIG.

This system is based on the system configuration of the homepage search system in the first embodiment. The difference is that the first keyword extracting means 213 provided in the center 200 in the first embodiment is connected to the terminal. This is the point of rearrangement on the 100 side. The other configurations are the same.

In the flowchart showing the processing flow of each means in the terminal 100 and the center 200,
Processing step S2 in FIG. 5 described in the first embodiment
25, the first keyword extraction processing step is eliminated. Therefore, in FIG. 6 showing an example of the data of the first search result in the center 200, the keyword of FstResult [] is not required. Instead, the first keyword extracting means 1 immediately before starting the processing step S121 in FIG. 9 described in the first embodiment, that is, immediately after the startup.
23 adds an execution step of converting (extracting) a keyword from text data of a homepage using the word dictionary 125. Accordingly, the storage device 103
Keyword to FstResult [] stored in
rd is added, and the converted (extracted) keyword data corresponding to the rd is stored.

As described above, by implementing the configuration, functions, and processing of the present embodiment, the conventional homepage search system could not obtain: In addition to further shortening the communication time (Internet connection time), Searching (narrowing down) is possible.-Since the information on the homepage is not provided to the user as it is, copyright issues are reduced.-Searching in natural language (users enter keywords in their own words). In addition, the processing load on the terminal side increases, but the load on the center side decreases, and a form of distributed processing can be realized. For example, software (program) In the case where it is necessary to change the keyword, a merit can be obtained in terms of cost. ・ The storage area for keywords in the storage device on the center side is no longer necessary. Therefore, the cost performance of the storage device is improved, and only one word dictionary used for keyword extraction is required.

The "first search ..." described in the first and second embodiments may be interpreted as "first search ..." or "first search ...". Similarly, “second search...”, “First keyword...”, “Second keyword...” May be interpreted in the same meaning as described above.

The contents of the above-described embodiments (the first and second examples) are not limited to the contents described above unless the gist of the present invention is changed.

[0109]

According to the homepage search system of the present invention, the following effects can be obtained in each claim.

According to the first aspect of the present invention, since the search can be performed step by step on the center side and the terminal side, the desired information can be obtained more accurately by the user. The user can input a keyword using his / her own words), so that the user can easily use it.

According to the second, third and fourth aspects of the present invention, the first search on the center side narrows down to the keyword level considerably, so that the communication time (internet connection time) can be further reduced. Further search (refinement) in the terminal becomes possible, and since the information on the homepage is not provided to the user as it is, there is an effect that copyright problems are reduced.

According to the fifth, sixth and seventh aspects of the present invention, the communication time (internet connection time) is narrowed down to the text level in the first search on the center side.
In addition to the further shortening of the URL, further search (refinement) in the terminal becomes possible, and the information on the homepage is not provided to the user as it is, so copyright issues are reduced,
Furthermore, the processing load on the terminal side increases, but on the other hand, the load on the center side decreases, and a storage area for keywords in the storage device on the center side is not required, thereby increasing the cost performance of the storage device and increasing the cost performance of keyword extraction. It is possible to obtain an effect that only one word dictionary is used.

[0113] In claim 8 of the present invention, the terminal side:
This has the effect that it is possible to reliably obtain the timing at which the search processing at the center is completed.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a system configuration of a homepage search system according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing an operation of a related URL search unit in the first embodiment of the present invention.

FIG. 3 is a display example of page data when a search service is disclosed on the Internet in the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a structure and contents of a URL list which is output contents of a related URL search means in the first embodiment of the present invention.

FIG. 5 is a flowchart showing an operation for generating a first search result which is a processing result of the first search means in the first embodiment of the present invention.

FIG. 6 is an explanatory diagram showing a structure and contents of a first search result delivered to a user as a processing result of a first search means in the first embodiment of the present invention.

FIG. 7 is an explanatory diagram showing an example of data downloaded and accessed by a search result processing means in the first embodiment of the present invention.

FIG. 8 is a flowchart showing the operation of the control means in the first embodiment of the present invention.

FIG. 9 is a flowchart showing an operation of a second search unit in the first embodiment of the present invention.

FIG. 10 is a display example of a second search screen according to the first embodiment of the present invention.

FIG. 11 is a functional block diagram showing a system configuration of a homepage search system according to a second embodiment of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 terminal 101 terminal display device 102 terminal input device 103 terminal storage device 104 terminal communication device 110 terminal control means 120 terminal second search means 200 center 201 center communication device 202 center storage device 210 center 1 Search means

Claims (8)

[Claims] 1. A terminal is connected to a center by a communication line, and by transmitting search data from the terminal to the center, a desired homepage is searched for on the center side. A homepage search center for transmitting and receiving data to and from the terminal and connecting to the Internet; and a search transmitted from the terminal. Storage means for storing search data and search results, receiving search data from the terminal via the communication means, and using a home page address related to the search data and a text portion in the home page as search results. Extracting, storing the search result in the storage means, and storing the primary search result in the terminal. The terminal is a homepage search terminal, a communication unit for transmitting and receiving data to and from the center, and a user required for search. Input means for inputting data; display means for displaying data or search results input by the input means; storage means for storing data and search results input by the input means; Transmitting the data for the primary search at the center input by the means to the center via the communication means, receiving the search result as the primary search result via the communication means, And control means for storing the first information from the center stored in the storage means.
Second search means for narrowing down the next search result using the data for the second search input by the input means and extracting the second search result, and performing stepwise homepage search. A featured homepage search system. 2. The method according to claim 1, wherein the first search means includes a related URL search means for searching a homepage address related to search data from the terminal, and a search result processing means for extracting a text portion in the homepage of the homepage address. , A word dictionary, and first keyword extracting means for converting the text portion into a plurality of keyword data using the word dictionary. 2. The homepage search system according to claim 1, wherein the search result is output in a search result. 3. The second detecting means, comprising: a word dictionary; and a second search means for converting the data for secondary search input from the input means into a plurality of keyword data using the word dictionary.
Keyword extraction means, and converts the keyword data of the primary search result by the first keyword extraction means and the output result by the second keyword extraction means into word vectors, performs a comparison process, and determines the degree of coincidence. 2. The homepage search system according to claim 1, wherein: 4. A primary search result output from said first search means is at least a homepage address U
3. The home page search system according to claim 2, comprising an RL character string, a title of the page, a text summary of the page, and a keyword included in the page text summary. 5. The first search means includes: a related URL search means for searching a homepage address related to search data from the terminal; and a search result processing means for extracting a text part in the homepage of the homepage address. 2. The homepage search system according to claim 1, further comprising: outputting the extracted text as a first search result. 6. The first detection means for converting a word dictionary and a text part of a primary search result from a center stored in the storage means into a plurality of keyword data using the word dictionary. Keyword extracting means, and second keyword extracting means for converting secondary search data input from the input means into a plurality of keyword data using the word dictionary, wherein the first keyword extracting means 2. The homepage search system according to claim 1, wherein the keyword data as the output result of the second keyword extraction unit and the keyword data as the output result of the second keyword extraction unit are respectively converted into word vectors, subjected to comparison processing, and the degree of coincidence is determined. . 7. A primary search result output from said first search means is at least a homepage address U
6. The home page search system according to claim 5, comprising an RL character string, a title of the page, and a text summary of the page. 8. The method according to claim 1, wherein the first search unit notifies the mail address received from the terminal that the primary search at the center has been completed when the primary search is completed. A homepage search system according to claim 1, 2 or 5.