U.S. patent application number 11/068208 was filed with the patent office on 2005-07-07 for document information search apparatus and method and recording medium storing document information search program therein.
This patent application is currently assigned to Fujitsu Limited. Invention is credited to Abe, Seiichiro.
Application Number | 20050149519 11/068208 |
Document ID | / |
Family ID | 18660740 |
Filed Date | 2005-07-07 |
United States Patent
Application |
20050149519 |
Kind Code |
A1 |
Abe, Seiichiro |
July 7, 2005 |
Document information search apparatus and method and recording
medium storing document information search program therein
Abstract
An apparatus for searching document information in response to a
search request from a client is disclosed. When a document file is
designated as a search condition by a search condition designating
unit of the client, the unit causes contents of the designated file
to be transmitted via a network. A document search unit of a search
machine forms a keyword based on the file contents transmitted from
the search condition designating unit, and searches similar
documents from an index or selection of important words extracted
from search target documents provided in a search database.
Inventors: |
Abe, Seiichiro; (Kawasaki,
JP) |
Correspondence
Address: |
Patrick G. Burns, Esq.
GREER, BURNS & CRAIN, LTD.
Suite 2500
300 South Wacker Drive
Chicago
IL
60606
US
|
Assignee: |
Fujitsu Limited
|
Family ID: |
18660740 |
Appl. No.: |
11/068208 |
Filed: |
February 28, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11068208 |
Feb 28, 2005 |
|
|
|
09761222 |
Jan 16, 2001 |
|
|
|
6883001 |
|
|
|
|
Current U.S.
Class: |
1/1 ;
707/999.005; 707/E17.058; 707/E17.109 |
Current CPC
Class: |
G06F 16/313 20190101;
G06F 16/38 20190101; Y10S 707/99936 20130101; Y10S 707/99933
20130101; G06F 16/30 20190101; G06F 16/9535 20190101 |
Class at
Publication: |
707/005 |
International
Class: |
G06F 017/30 |
Foreign Application Data
Date |
Code |
Application Number |
May 26, 2000 |
JP |
2000-155867 |
Claims
1-9. (canceled)
10. A document information search apparatus comprising: a database
in which index information describing a list of important words
extracted from search target documents has been stored every
document; a text extraction processing unit which extracts a text
document from contents of a file received in response to a search
request from a network in which a document file has been designated
as a search condition; a morpheme analyzing unit which extracts
nouns by a morpheme analysis of said text document; a keyword
forming unit which extracts important words from said nouns and
forms a keyword in which said important words are coupled by OR;
and a search executing unit which searches similar documents by
searching said database by said keyword and notifies a requesting
source of a search result.
11. An apparatus according to claim 10, wherein said keyword
forming unit counts the number of times of appearance showing in
which documents in the index of each of the search documents stored
in said document database each of said nouns appears, selects a
predetermined number of upper words each having the number of times
of appearance in a predetermined range, and forms the keyword.
12. An apparatus according to claim 10, wherein property
information extracted from the search target documents is stored in
said database together with the index information, said keyword
forming unit allows said property information extracted from the
file received in response to the search request to be included in
said keyword, thereby allowing the similar documents to be
searched.
13. A document information search method of searching document
information on the basis of a search request transmitted via a
network and responding, comprising the steps of: storing index
information describing a list of important words extracted from
search target documents every document into a database; in the case
where a file is designated as a search condition on a search
requesting source, transmitting contents of the designated file to
a server together with the search request through the network; and
on a search side, extracting a text document from the file contents
received in response to the search request, extracting nouns by a
morpheme analysis of said text document, extracting important words
from said nouns, forming a keyword in which said important words
are coupled by OR, searching similar documents by searching said
database by said keyword, and responding a search result.
14. A method according to claim 13, wherein as a formation of said
keyword, the number of times of appearance showing in which
documents in the index of each of the documents stored in said
database each of said nouns appears is counted, a predetermined
number of upper words each having the number of times of appearance
in a predetermined range are selected, and the keyword is
formed.
15. A method according to claim 14, wherein property information
extracted from the file received in response to the search request
is included in said keyword and the similar documents are
searched.
16-18. (canceled)
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The invention relates to document information search
apparatus and method for promptly searching necessary documents
from a large amount of document data and to a recording medium in
which a document information search program has been stored. More
particularly, the invention relates to document information search
apparatus and method for searching documents having similar
contents by a simple operation such that a document file itself is
designated as a search condition and to a recording medium in which
a document information search program has been stored.
[0003] 2. Description of the Related Arts
[0004] Hitherto, in a document management system using a network
environment, a document information search apparatus which can
search necessary documents from a large amount of document data
existing on the Internet or Ethernet and can rapidly refer to it
has been provided. In a document search in such a case, the user
designates one or a plurality of proper words or character trains
which are considered to be included in the necessary documents as
keywords and searches the documents including the word(s) of the
designated keyword from a search database, and a document list is
displayed as a search result. In such a document information search
apparatus, with respect to search target documents existing on the
network, an index describing a list of important words extracted
from their contents is formed every document and stored in the
search database. If a search request in which a keyword is
designated is issued from the user, the index in the search
database is searched, thereby outputting a search result of the
document list. Further, according to the conventional document
information search apparatus, after the user searched the documents
which are considered to be necessary from the document list
searched by the designation of the keyword, if a search of similar
documents is designated with respect to the selected documents,
words in which a frequency of appearance in the search documents is
high are automatically extracted and added to previously executed
search conditions on the basis of a condition of the OR, so that
the similar documents can be searched.
[0005] However, in the case where the user wants to obtain an
interesting document and search documents with contents similar to
such a document via Email or the Internet, in the present
situation, a word or character train included in the obtained
document is selected and designated as a keyword each time and
documents including the word of the designated keyword are searched
from the database, thereby obtaining a document list. Subsequently,
the user has to select a document from the searched document list,
designate a similar document search, and search similar documents
from the database. That is, even if the user desires to search
similar documents of the document obtained by the Email or
Internet, according to the conventional document information search
apparatus, with respect to only the document which has already been
registered in the search database, the similar documents can be
searched by using such a registered document as a search condition,
and the similar documents cannot be directly searched by using the
document obtained by the user by the Email or Internet as a search
condition.
[0006] When the user selects keywords which are considered to be
necessary for document search from the document obtained by the
Email or Internet and inputs such keywords as a search condition,
if there are many keywords, it is troublesome to input them. Unless
the keywords are sufficiently designated, there is a case where a
search leakage occurs and an expected search result is not
obtained. Further, there is also a case where the number of search
documents which are obtained as a document list becomes extremely
large. There is a case where it is very laborious to search
necessary documents by opening documents which are considered to be
concerned with target documents from the document list.
SUMMARY OF THE INVENTION
[0007] According to the invention, document information search
apparatus and method in which documents similar to a document which
is not registered in a search database can be promptly searched by
a simple operation are provided and a computer-readable recording
medium in which a document information search program has been
stored is provided.
[0008] According to the invention, there is provided a document
information search apparatus which searches document information
and responds on a search side such as a server or the like on the
basis of a search request sent from a client or the like via a
network, wherein: a search condition designating unit which, when a
file is designated as a search condition, transmits contents of the
designated file through the network is provided for a requesting
source such as a client or the like; and a search machine which
forms a keyword from the file contents transmitted from the search
condition designating unit and searches similar documents is
provided on a search side such as a server or the like. Therefore,
if the user wants to obtain documents including interesting
contents by Email, Internet, or the like and search documents of
contents similar to such a document or the like, a file which was
uploaded by a designation of a document is designated as a search
condition, thereby enabling the documents having similar contents
to be searched. Therefore, any document which is not registered in
a database can be freely designated as a search condition, a
troublesome input of keywords based on the document contents
becomes unnecessary, and the similar documents can be easily and
promptly searched.
[0009] The search condition designating unit on the search
requesting source transmits a head file portion of the designated
file contents. Since many important keywords necessary for document
search exist usually in a head portion of a document, only the head
portion of the file contents, for example, the head portion of 1 kB
is transmitted as a search condition. Since the document files
which are used for the search condition have various sizes, by
deciding a capacity of the file which is transmitted as a search
condition, a communication load and the processes on the search
side are reduced. The search condition designating unit includes an
HTML file and an Excel file as files which are designated as a
search condition. Even in the other file formats, the files include
a file of an arbitrary file format so long as it is a file from
which a text document can be extracted. A database in which index
information describing a list of important words extracted from the
search target documents has been stored every document is provided
for the search machine on the server side. The document search unit
of the search machine comprises: a text extraction processing unit
which extracts a text document from the file contents received in
response to the search request; a morpheme analyzing unit which
extracts nouns by a morpheme analysis of the text document; a
keyword forming unit which extracts important words from the nouns
and forms a keyword in which the important words are coupled by OR;
and a search executing unit which searches similar documents by
searching the search database by the keyword and notifies the
client of a search result. The keyword forming unit counts the
number (H) of times of appearance showing in which document in the
index of each search document stored in the search database each
noun appears and selects a predetermined number of upper words each
having the number (H) of times of appearance in a predetermined
range, thereby forming the keyword. When the number of documents in
the index is assumed to be N, the keyword forming unit selects
upper ten words each having the number of times of appearance in a
range in which the number (H) of times of appearance is equal to,
for example,
2N/3.gtoreq.H.gtoreq.1
[0010] and forms the keyword. Thus, the important words which have
been registered in the index of the database and are necessary for
the similar document search of the existing documents are narrowed
down, thereby raising precision of the similar document search.
Further, the keyword forming unit allows property information
extracted from the received file to be included in the keyword and
searched in response to the search request. As property information
in this case, there are a writer of the file received in response
to the search request, a document title, and the like. By adding
the property information of the file into the search conditions as
mentioned above, for example, in the case where the user wants to
specify the writer or the like, the similar documents can be
properly narrowed down. The search condition designating unit of
the search requesting source is provided by a WWW browser of the
client and the contents of the file designated on a search request
picture plane of the WWW browser are transmitted to a WWW server
through the network and sent to the search machine.
[0011] The invention also provides the search machine serving as a
document information search apparatus on the search side such as a
server or the like. The document information search apparatus
serving as a search machine comprises: a search database in which
index information describing a list of important words extracted
from search target documents has been stored every document; a text
extraction processing unit which extracts a text document from the
contents of the file received by a search request which is
transmitted from a network in which a document file has been
designated as a search condition; a morpheme analyzing unit which
extracts nouns by a morpheme analysis of the text document; a
keyword forming unit which extracts important words from the nouns
and forms a keyword in which the important words are coupled by OR;
and a search executing unit which searches similar documents by
searching the search database by the keyword and notifies a
requesting source of a search result.
[0012] The invention provides a document information search method
of searching document information and responding on a search
machine side such as a server or the like on the basis of a search
request which is transmitted from a search requesting source such
as a client or the like via a network, comprising the steps of:
storing index information describing a list of important words
extracted from search target documents into a search database of
the server every document; when a document file is designated as a
search condition, transmitting contents of a designated file to a
search side via the network together with the search request; and
on the search side, extracting a text document from the received
file contents in response to the search request, extracting nouns
by a morpheme analysis of the text document, extracting important
words from the nouns, forming a keyword in which the important
words are coupled by OR, searching similar documents by searching a
search database by the keyword, and notifying the client of a
search result. The details of the document information search
method are fundamentally the same as those of the apparatus
construction.
[0013] Further, the invention provides a computer-readable
recording medium in which a document information search program has
been stored, wherein the document information search program
comprises the steps of: receiving a search request in which a
document file is designated as a search condition; extracting a
text document from contents of the file received in response to the
search request; extracting nouns by a morpheme analysis of the text
document; extracting important words from nouns and forming a
keyword in which the important words are coupled by OR; and
searching similar documents by searching a database by the keyword
and notifying a requesting source of a search result.
[0014] The above and other objects, features, and advantages of the
present invention will become more apparent from the following
detailed description with reference to the drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] FIG. 1 is an explanatory diagram of a system construction of
the invention;
[0016] FIG. 2 is a block diagram of a functional construction of
the invention;
[0017] FIG. 3 is a block diagram of a document search unit
according to the invention;
[0018] FIG. 4 is an explanatory diagram of a process of a search
database forming unit in FIG. 2;
[0019] FIG. 5 is a flowchart for a browser process in FIG. 2;
[0020] FIGS. 6A and 6B are explanatory diagrams for the search
requesting operation in which a document file is designated as a
search condition according to the invention;
[0021] FIG. 7 is a flowchart for a server searching process of the
invention;
[0022] FIG. 8 is a flowchart for a text extracting process in FIG.
7;
[0023] FIGS. 9A and 9B are explanatory diagrams for extracting a
text document from an HTML file by the process in FIG. 7;
[0024] FIG. 10 is an explanatory diagram of property information
provided in the HTML file which is used for search according to the
invention;
[0025] FIGS. 11A and 11B are explanatory diagrams of an Excel
document which is used as a text extraction target by the process
in FIG. 7; and
[0026] FIG. 12 is an explanatory diagram of a text document
extracted from the Excel document in FIGS. 11A and 11B.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
[0027] FIG. 1 shows a system construction of a document information
search apparatus according to the invention and relates to an
example in the case where the system is constructed as a search
system of a server client type using the Internet or Ethernet. A
client 12 on the user side is connected to a server 10 via an
Internet/Intranet 14. A WWW browser 16 for search is provided for
the client 12. A search request of document information is issued
to the server 10 by using the WWW browser 16 and a search result on
the server 10 side is displayed. A WWW server 18, a search machine
20, and a document database 24 are provided for the server 10. A
search database 22 has been stored in the search machine 20. Search
target documents 25 have been stored in the document database 24.
Further, external document management servers 44 and 48 are
connected to the WWW server 18. Document databases 46 and 50 are
also provided for the document management servers 44 and 48. The
search target documents 25 have been stored in the document
databases 46 and 50, respectively. The WWW server 18 provided for
the server 10 receives the search request from the browser 16 and
requests the search machine 20 to search. The search result
returned from the search machine 20 is sent to the browser 16 and
the browser 16 is allowed to display it. To process a whole
sentence search to be processed at a high speed, the search
database 22 functions as a storage for managing an index formed by
a set of important words described in a document serving as a
search target. A document name of the document and its storing
location are recorded in the index. When the search request is
received from the browser 16, the search machine 20 executes a
searching process with respect to the index in the search database
22 as a target. The search target documents 25 collected from the
document management servers 44 and 48 have been stored in the
document database 24. The index in the search database 22 is formed
with respect to the document database search target documents 25 as
targets. In such a search system of the server client type, a
search condition designated by the user is sent to the WWW server
18 on the server 10 side via the Internet/Intranet 14 by using the
browser 16 of the client 12. The designated search condition
included in the search request received by the WWW server 18 is
sent from the WWW server 18 to the search machine 20. The search
machine 20 searches the documents which conform with the search
condition from the search database 22 and notifies the WWW server
18 of a search result. The WWW server 18 sends the search result
from the search machine 20 to the browser 16 of the client 12 and
allows the browser 16 to display it. The user looks at the search
result processed by the browser 16 and selects a link described in
the search result, so that he can upload his desired search target
documents 25 from the selected documents via the WWW server 18 and
see the contents of the desired document 25.
[0028] FIG. 2 is a block diagram of a functional construction in
the search system of FIG. 1. First, a search condition designating
unit 26 is provided for the WWW browser 16 serving as a user side.
The search condition designating unit 26 of the invention directly
designates a document file, as a search condition, obtained by the
user as a search condition via the Internet, Email, or the like,
and transmits the contents of the designated file to a document
search unit 30 of the search machine 20 via the WWW server 18
through the Internet/Intranet 14. Besides the search condition of
the file designation which is newly provided in the invention, the
search condition designating unit 26 can also designate the
following search conditions.
[0029] I. Keyword search
[0030] II. Detail search in which a keyword is designated for each
of a title of the document, a writer, and a text, and a search is
performed.
[0031] III. Sentence search in which by inputting a daily word or
sentence, the documents concerned with the text contents are
searched.
[0032] IV. Similar document search in which the existing document
registered in the search database 22 is used as a search
condition.
[0033] And the like.
[0034] A search database forming unit 28, the document search unit
30, and a document reference unit 32 are provided for the search
machine 20 provided on the WWW server 18 side. The search database
forming unit 28 forms an index in the search database 22 and
registers it. That is, with respect to each of the search target
documents 25 collected and stored in the document database 24, the
search database forming unit 28 extracts important words described
in the search target documents 25, forms an index constructed by a
set of extracted words, and stores it. Naturally, the document
name, storing location, and the like of the search target documents
have also been recorded in the index. The document search unit 30
forms a keyword from the file contents at the time when the file is
designated as a search condition transmitted from the search
condition designating unit 26 in the WWW browser 16, performs a
search collation with the set of important words included in the
index in the search database 22, searches documents similar to the
document of the file designated as a search condition by the WWW
browser 16, returns a search result from the WWW server 18 to the
WWW browser 16, and allows the WWW browser 16 to display it. When
the document to be referred to is selected from a document list as
a search result sent by the WWW browser 16 and notified to the
document reference unit 32 through the WWW server 18, the document
reference unit 32 extracts a requested reference document from the
document database 24 and returns it to the WWW browser 16.
[0035] FIG. 3 shows the details of a functional construction of the
document search unit 30 of the invention provided for the search
machine 20 in FIG. 2. A search designation file storing unit 34, a
text extraction processing unit 36, a morpheme analyzing unit 38, a
keyword forming unit 40, and a search executing unit 42 are
provided for the document search unit 30. An index 52 comprising a
set of important words, the document name, storing location, and
the like of each of the search target documents 25 in the document
database 24 formed by the search database forming unit 28 in FIG. 2
has been stored in the search database 22. The file contents
transmitted by the file designation of the search condition
designating unit 26 in the WWW browser 16 in FIG. 2 are stored in
the search designation file storing unit 34 in the document search
unit 30. When the file contents are transferred from the WWW
browser 16 side, a head file portion of the document file
designated as a search condition, for example, 1 kB of the head
portion is extracted and transmitted together with the search
request to the WWW server 18 side. A capacity of the file which is
transmitted as a search condition is set to a fixed capacity, for
example, 1 kB as mentioned above, thereby setting a transfer load
of the document contents to the search machine 20 side to be
constant irrespective of a size of document file designated as a
search condition. The searching process by the document search unit
30 in the search machine 20 is stabilized and a high processing
speed is realized. The text extraction processing unit 36 extracts
a text document from the file contents designated as a search
condition stored in the search designation file storing unit 34. As
a format of the document file which is designated as a search
condition in the WWW browser 16, there are various file formats
such as text file of Email, HTML file in the Internet, further,
Excel file of an aggregate list, and the like. Therefore, to enable
a search function to be presented with respect to a difference of
the file formats, only the text document is extracted from the
document files of various formats by the text extraction processing
unit 36 and used as a search condition. The morpheme analyzing unit
38 subsequently provided extracts nouns included in the extracted
text document by using a morpheme analysis. The nouns in the
document contents extracted by the morpheme analyzing unit 38 are
sent to the keyword forming unit 40. The keyword forming unit 40
extracts important nouns in order to form a keyword. As for the
extraction of the important words in the keyword forming unit 40,
first, the number (H) of times of appearance showing in which
documents in the number (N) of documents registered in the index 52
in the search database 22 each noun appears is counted. When the
number (H) of times of appearance of document in the index 52 is
obtained, words in which the number (H) of times of appearance lies
within a predetermined range, for example,
(2N/3).gtoreq.H.gtoreq.1
[0036] are selected. Upper ten words in which the number (H) of
times of appearance is large among the words selected as mentioned
above are selected to form the keyword. A query expression in which
the 10 selected important words are coupled by OR is formed and
provided to the search executing unit 42. On the basis of the query
expression derived from the keyword forming unit 40, the search
executing unit 42 performs a search collation with the index 52 in
the search database 22, extracts an index which satisfies a
predetermined similarity as a search result, and transmits the
search result to the WWW browser 16 side by the WWW server 18,
thereby enabling the user to refer to the search result in a form
of a document list. Further, the document search unit 30 can also
perform a document search using property information of the file
designated as a search condition stored in the search designation
file storing unit 34. For this purpose, when the document file is
designated as a search condition, the search condition designating
unit 26 in the WWW browser 16 extracts the property information of
the designated document file and transmits the property information
to the search machine 20 side together with the head file portion,
for example, 1 kB of the head file portion of the document
designated as a search condition. In the document search unit 30 in
FIG. 3, in addition to the extraction of the text document from the
file contents, the extraction of the nouns by the morpheme
analysis, and the formation of the keyword by the selection of the
important words with respect to the nouns, for example, a date of
formation, a writer, a title, and the like are extracted from the
property information added to the file contents stored in the
search designation file storing unit 34. The property information
is included in the keyword by the keyword forming unit 40. The
index 52 in the search database 22 is searched by the search
executing unit 42.
[0037] FIG. 4 is an explanatory diagram of the index forming
process by the search database forming unit 28 provided in the
search machine 20 in FIG. 2. In the search database forming unit
28, a robot 54 collects documents 66 from the external document
databases 46 and 50 and stores them into a temporary file 62 and,
at the same time, adds a list of the collected documents 66 into a
collection document list file 64. The robot 54 subsequently sends a
process to a text extracting unit 56. The text extracting unit 56
extracts the collected documents 66 from the collection document
list file 64 and stores them into an extraction text file 68.
Subsequently, the process is sent to an important word extracting
unit 58. The important word extracting unit 58 extracts nouns from
the relevant text document in the extraction text file 68 by a
morpheme analysis, counts the number of times of appearance with
respect to each noun, extracts, for example, upper ten words in
which the number of times of appearance is high as important words,
and stores them into an important word file 70. Subsequently, the
process is sent to an index forming unit 60. The index forming unit
60 extracts a set of, for example, upper ten important words with
respect to the relevant document, further forms an index in which
the document name and the storing location have been added, and
stores it as index information into the search database 22.
[0038] FIG. 5 is a flowchart for a browser process for performing
the designation of the search condition and the display of the
search result by the WWW browser 16 in FIG. 2. When the user opens
the search function of the WWW browser 16, a search picture plane
is displayed in step S1. When the search picture plane is
displayed, a designating operation of the search condition in which
the document file has been designated is performed in step S2.
Subsequently, in step S3, whether the search has been activated or
not is discriminated. When the search activation is determined,
whether the search is a file designating search or not is
discriminated in step S4. If YES, step S5 follows and the file
designated by the user is read out. In step S6, 1 kB of the head in
the designated file is transmitted to the server together with the
search requesting message. If the search is not the file
designating search, a search requesting message corresponding to
the other search, for example, a keyword search is transmitted to
the server in step S7. When the head portion in the designated file
is transmitted to the server in step S6, the apparatus waits for
reception of the search result in step S8. When the search result
is received from the server in step S8, step S9 follows and the
user executes a display operating process of the search result and
looks at the search contents. Such processes in steps S1 to S9 are
repeated until a search end instruction for closing the search
picture plane is issued in step S10.
[0039] FIGS. 6A and 6B show a specific procedure and a state of the
picture plane in the case where the document file is designated as
a search condition in the browser process in FIG. 5. First, the
user obtains a document file 72 to be designated as a search
condition from, for example, the Internet. The user looks at the
contents in the document file 72 and stores the contents in the
document file 72 into a predesignated file, for example, a file
"news.txt" in order to search documents similar to the document
file 72. The user subsequently opens a keyword input picture plane
74. A keyword input unit 76, a file designating unit 78, a Refer
button 80, and an Execute search button 82 are provided for the
keyword input picture plane 74. The user presses the Refer button
80 on the keyword input picture plane 74, thereby displaying a file
selection dialog 84. Since the document file 72 which the user
wants to designate as a search condition has been stored in the
file selection dialog 84, by selecting the file name "news.txt" by
clicking a mouse, the selected file name "news.txt" is set into the
file designating unit 78 on the keyword input picture plane 74.
After completion of the file designation by the file designating
unit 78 as mentioned above, by pressing the Execute search button
82, 1 kB of the head portion of the document contents in the
document file "news.txt" designated as a search condition is
transmitted to the server together with the search request.
[0040] FIG. 7 is a flowchart for the server searching process which
is realized by the document search unit 30 in FIG. 3. In the server
searching process, the document file designated as a search
condition is read in step S1. A process for extracting the text
document from the document file is performed in step S2.
Subsequently, nouns are extracted from the contents of the
extracted text document by using the morpheme analysis in step S3.
A process for counting the number (H) of times of appearance
showing in which documents in the number (N) of documents in the
index 52 provided for the search database 22 each word extracted as
a noun appears is performed in step S4. If the number (H) of times
of, appearance in the index of each word can be counted, in step
S5, the words in a range in which the number (H) of times of
appearance is equal to or smaller than (2N/3) and is equal to or
larger than 1 (1.ltoreq.H.ltoreq.2N/3) are first selected and upper
ten words in which the number (H) of times of appearance is large
among the selected words are selected as important words which are
used as a keyword. Subsequently, a query expression in which the 10
words selected as important words are coupled by OR is formed in
step S6. In step S7, the index in the search database by the query
expression formed as a search keyword is searched and the contents
of the index having a predetermined similarity for the formed
keyword are collected as search documents into a list table. The
search result is transmitted to the browser in step S8.
[0041] FIG. 8 shows the details of the text extracting process in
step S2 in FIG. 7. In the text extracting process, an extension of
the document file is decoded in step S1. When it is recognized from
the file extension that the text is an HTML document in step S2,
step S3 follows. Data in a body tag in the HTML document is
extracted as a text data main body and tag data is removed. For
example, in case of an HTML file as shown in FIG. 9A, data in a
body word surrounded by < > is extracted as a text data main
body and the tag data is removed, so that an extraction text
document as shown in FIG. 9B is obtained. Subsequently, property
information of the file managed by the OS is obtained in step S4.
The property information includes, for example, a file owner, a
document type, and the like.
[0042] FIG. 10 shows an example of the property information of the
document file obtained from the Internet. In this property
information, a document title "With respect to document management
system", a date of formation, a date of change, and the like exist.
Those property data is obtained to form a keyword. If the document
is not the HTML document but, for example, an Excel document or the
like in step S2, the file is sent to a document library and text
data is obtained in step S5. Subsequently, in step S6, the file
property information set every document, for example, the writer,
the document title, and the like are obtained by a property
information obtaining function.
[0043] FIGS. 11A and 11B show an Excel file as a file (other than
the HTML file) which is designated as a search condition in the
invention. With respect to this Excel file, when it is sent to the
document library and the text data is obtained, a result obtained
by extracting the text document written in the Excel document as
shown in an extraction text document in FIG. 12 is obtained. The
text documents from the HTML document and Excel document obtained
by such a text extracting process and, further, the text documents
derived from the property information are collected and nouns are
extracted by using the morpheme analysis in step S3 in FIG. 7. In
steps S4 and S5, upper ten important words are selected as keywords
with reference to the index in the database, a query expression is
formed, the index in the database is searched, and a search result
can be obtained. When obtaining the property information in steps
S4 and S6 in the text extracting process in FIG. 8, whether the
property information is used or not can be selected by the
designation on the user side in the WWW browser 16. The
discrimination about the use of the property information depends on
a judgment of the user to decide to which extent the search result
is narrowed down.
[0044] The invention also provides a computer-readable recording
medium in which a document information search program for allowing
the search machine 20 in FIG. 3 to execute a processing function of
the document search unit 30 has been recorded. An embodiment of the
recording medium incorporates a removable portable recording medium
such as CD-ROM, floppy disk, or the like, a storing device of a
program provider who provides the program by a line, and further, a
memory device such as RAM, hard disk, or the like of a processing
apparatus in which the program has been installed. The document
information search program which is provided by the recording
medium and used for realizing the functions of the document search
unit 30 in FIG. 3, specifically speaking, the document information
search program having processing steps of executing the processes
shown in the flowcharts of FIGS. 8 and 9 is installed into the
processing apparatus such as a server or the like and executed on a
main memory. When a service request is received from the client
side, the document information search program of the invention
installed to the server side uploads the WWW browser function for
designating the search condition by the file designation onto the
client 12 side, thereby enabling the user to use the search
system.
[0045] According to the invention as mentioned above, when the user
obtains the document including the interesting contents by the
Email, Internet, or the like, in the document search of the
contents similar to such a document, by directly designating the
document file as a search condition, the documents having similar
contents can be simply and rapidly searched. The troublesome
inputting operation of the keyword based on the document contents
becomes unnecessary. The search of the similar documents by the
user can be extremely efficiently realized. In the formation of the
keyword necessary for the document search by the file designation,
besides the operation for extracting the important words from the
document contents, the important words are extracted also from the
property information held in the document file and included in the
keyword, so that the similar search of the existing documents
registered in the database is more properly narrowed down.
Precision of the search can be raised.
[0046] Although the embodiment has been described with respect to
the example of the search system of the server client type, the
invention is not limited to it but many other system forms such as
a host terminal type can be used. The invention is not limited to
the foregoing embodiment but incorporates many proper modifications
and variations without losing the objects and advantages of the
invention. Further, the invention is not limited by the numerical
values shown in the embodiment.
* * * * *