JP2003044493A - Patent information searching device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a patent information searching device which supports a search effectively utilizing a code. SOLUTION: Concerning the set of patent documents searched from a database 3, in a control part 11, a classification code applied to each of patent documents is acquired and a higher-order code part is extracted from the relevant classification code. Concerning the higher-order code, the frequency of appearance in the set of searched patent documents is operated, display blocks corresponding to respective higher-order codes are formed in the order of the relevant frequency, a tendency analytic table is generated by arraying the classification codes belonging to the correspondent higher-order codes in the order of classification codes within that display block, and such a table is provided to a user. Besides, the patent information searching device is provided with a tendency analytic image for mapping the classification code applied to the searched result on a classification code explanation statement image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a patent information retrieval device for retrieving patent documents, and more particularly to improvement of selection processing of classification codes used for retrieval.

[0002]

2. Description of the Related Art Many patent documents are cumulatively accumulated year by year. Such patent documents will be published early before the acquisition of rights ("Patent Publication" in Japan).
(A)) and those publicly announced at the time of patenting (“patent publication” (B2) in Japan). The latter is used to indicate the scope of rights and is used for purposes such as avoidance of infringement and technical cooperation, the former is used as technical literature, and in many countries, it serves as a basis to prevent patentability of subsequent patent applications. It is used for patent office work such as examination, and in any case, it is an extremely effective technical document. In the following description, the term "patent document" includes not only documents relating to ordinary patent applications but also utility model registration applications.

However, the amount thereof increases cumulatively year by year, and not only is it difficult to use, but also the burden of examination work is heavy. Therefore, the "international patent classification" code is used as an international agreement in order to classify the patent documents in advance by the technical contents and facilitate the search. This code
A "section" that roughly divides the technical field, a "subsection" obtained by dividing each section, a "class" that subdivides the technology belonging to the subsection, a "subclass" that subdivides the class, and a subclass that subdivides It is a hierarchical code composed of "groups" and "subgroups" which are more detailed classifications.

This code is, for example, "G06F 17/6
"G" is added like "0", the leading "G" means a section, the following "06" indicates a class, and the following "F" indicates a subclass. And "17/60" means a subgroup. It should be noted that the subgroup is a detailed version of a subgroup that belongs to the technical field of the main group, such as "17/00", in which "00" follows the slash. The code part up to the subclass is, for example, "G06
"F" is "electrical digital data processing", "G06G" is "analog computer" (according to International Patent Classification 7th Edition), and adjacent codes are conceptually orthogonal to each other, and indicate almost independent themes. On the other hand, "G06F 17
/ XX ”are all“ devices or methods for digital computation or data processing specially adapted for a particular function ”,
For example, "G06F 17/50" is CAD (Computer Utilization Design), and "G06F 17/60" indicates that a computer is used for "management purpose, business purpose, management purpose, supervisory purpose or prediction purpose". Similar technical contents are arranged between adjacent codes.

Other codes similar to the international patent classification include "FI" which is improved based on the international patent classification and "F term" which is a highly accurate classification mainly developed for examination by the Patent Office. , Etc. In the United States, a classification code called "US class" is used. In any of these, similar to the above-mentioned international patent classification, a part that represents a concept (theme) that is almost independent between adjacent codes, and a subordinate of that part that represents continuous technical contents between adjacent codes Comprising parts and.

Conventionally, most of the devices for searching patent documents have
Search these codes assigned to the patent document,
A process for searching for patent documents having a designated code is performed. Specifically, as a document showing such a device,
Japanese Patent Laid-Open No. 5-128152 (“Document Search Method”),
Japanese Patent Application Laid-Open No. 8-339380 (“Information Retrieval Processing Method”), Japanese Patent Application Laid-Open No. 11-203292 (“Data Map Creating Device and Recording Medium Recording Processing Program Used by the Device”), Japanese Patent Laid-Open No. 2000-322447.
JP-A No. 2001-52022 (“Data collection / reference type search system”) and the like.

[0007]

However, in the search device of the above-mentioned conventional patent documents, since the code is specified in detail in, for example, the F term, even if it is limited to a computer suitable for a specific application, for example, "5B" showing the theme
049 "(theme code; upper code) and below, an extremely large number of lower codes (F terms) belong, and it was extremely difficult for the user to select an F term that matches the search purpose. Although this situation is remarkable in the detailed classification such as the F term, the international patent classification and the FI,
Even in the US class, there were some similar problems. Further, if these classification codes are simply rank-processed in the order of the frequencies assigned to the document set for which the keyword search is performed (see, for example, the one described in JP 2001-52022 A), the discrete classification codes However, if the adjacent classification code is originally appropriate, the classification code is not presented, and the search that effectively uses the classification code cannot be sufficiently supported.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a patent information search device capable of supporting a search using a code effectively.

[0009]

DISCLOSURE OF THE INVENTION The present invention for solving the above-mentioned problems of the prior art is a patent information retrieval device, which is constructed from a database accumulating patent documents to which classification codes having a plurality of layers are added. A search means for searching for patent documents according to the input search conditions, and a tendency analysis showing the appearance tendency of the classification code assigned to each document for a document set consisting of at least one patent document obtained as a result of the search. An image generating means for generating an image, wherein the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, and the image generating means includes each upper code of the classification code and the upper code. A first display section for displaying the technical field information corresponding to the code in association with each other, each sub code of the classification code, and the technical field information corresponding to the sub code. A means for generating a commentary image including a second display portion additionally displayed; and a classification code assigned to the searched document, which corresponds to at least one of a higher code and a lower code, Means for emphasizing at least a part of the commentary image and thereby generating the tendency analysis image.

The present invention for solving the above-mentioned problems of the prior art is a patent information retrieval apparatus, which is based on a retrieval condition input from a database accumulating patent documents to which classification codes having a plurality of layers are added. , A search means for searching patent documents and a document set consisting of at least one patent document obtained as a result of the search are subjected to aggregation processing using the classification code, and a tendency analysis showing the aggregation processing result Aggregation processing means for creating a table, wherein the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, and the aggregation processing means sets a section for each of the upper codes, It is characterized by including means for displaying the lower order code in each section and creating the tendency analysis table by this.

The present invention for solving the above-mentioned problems of the prior art is a patent information retrieval apparatus, which is constructed in accordance with a retrieval condition input from a database accumulating patent documents to which classification codes having a plurality of layers are added. Image generation for generating a tendency analysis image representing a tendency of appearance of a classification code assigned to each document with respect to a document set including at least one patent document obtained as a result of the search. And a totalization processing means for performing a totalization process on a document set consisting of at least one patent document obtained as a result of the search, using the classification code, and creating a trend analysis table representing the totalization process result. When,
And a display unit for scrolling and displaying at least one of the trend analysis image and the trend analysis table, wherein the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, and the image generation unit. Is a first display section for displaying each higher code of the classification code and the technical field information corresponding to the higher code in association with each other, each lower code of the classification code and the technical field information corresponding to the lower code. A means for generating a commentary image including a second display portion that is displayed in association with each other; and a classification code assigned to the searched document, which corresponds to at least one of a higher code and a lower code, Means for emphasizing at least a part of the commentary image and thereby generating the tendency analysis image, wherein the totalization processing means includes the upper code. Set the compartment and to display the lower code in each compartment, thereby comprises means for creating the trend analysis table, said display means,
When the upper code or the lower code on the tendency analysis table is selected, the portion of the tendency analysis image corresponding to the upper code or the lower code is scroll-displayed.

[0012] Here, the totalizing means includes means for calculating a first appearance frequency for each higher-order code found in the patent document subset, and classification code found in the patent document subset. Means for calculating a second appearance frequency, the display order of the sections is adjusted on the basis of the first appearance frequency, and a sub-code corresponding to each classification code in each section is adjusted on the basis of the second appearance frequency. It is preferable that the display order is adjusted.

Further, the apparatus further comprises means for designating a classification code as a search condition on the tendency analysis image or the manual image, and the designated classification code is provided for the search processing as at least a part of the search condition. Is also preferable.

Further, the search means is selected by a preliminary search means for performing a preliminary search based on the input search conditions and an instruction to select a part of the documents obtained as a result of the preliminary search. And a means for outputting at least one patent document obtained as a result of the search, and at least one patent document output as the result of the search may be subjected to a predetermined process as the document set. preferable.

The present invention for solving the above-mentioned problems of the prior art is a patent information search method, wherein a patent is searched according to an input search condition from a database accumulating documents to which classification codes of a plurality of layers are added. A search step for searching documents and at least one obtained as a result of the search.
An image generation step of generating a tendency analysis image representing the appearance tendency of the classification code assigned to each document with respect to a document set consisting of two patent documents, wherein the classification code is a high-order code indicating a high-order classification and a low-order code. A lower-order code indicating a classification, and the image generation step includes a first display unit that displays each higher-order code of the classification code and technical field information corresponding to the higher-order code in association with each other, and a lower-order code of each classification code. A step of generating a commentary image including a second display unit that associates and displays a code and technical field information corresponding to the lower order code; and a higher order of the classification code assigned to the searched document. Emphasizing at least a part of the commentary image corresponding to at least one of a code and a subordinate code, thereby generating the tendency analysis image. It is characterized.

The present invention for solving the above-mentioned problems of the prior art is a method for searching patent information, which is based on a search condition input from a database accumulating patent documents to which classification codes having a plurality of layers are added. , A search step of searching for patent documents, and performing a totalization process using a classification code on a document set consisting of at least one patent document obtained as a result of the search, and a tendency analysis showing the totalization process result. A totalization processing step of creating a table, wherein the classification code includes a high-order code indicating a high-order classification and a low-order code indicating a low-order classification, and the calculation processing step sets a section for each of the high-order codes, It is characterized in that the subordinate code is displayed in each section and the trend analysis table is created by this.

The present invention for solving the above-mentioned problems of the prior art is a method for searching patent information, which is based on a search condition input from a database accumulating patent documents to which classification codes having a plurality of layers are added. Image generation for generating a tendency analysis image showing a tendency of appearance of a classification code assigned to each document with respect to a document collection consisting of at least one patent document obtained as a result of the search And a totalization process of performing a totalization process on a document set consisting of at least one patent document obtained as a result of the search using the classification code, and creating a trend analysis table representing the totalization process result. When,
A display step of scrolling and displaying at least one of the trend analysis image and the trend analysis table, wherein the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, and the image generating step. Is a first display section for displaying each higher code of the classification code and the technical field information corresponding to the higher code in association with each other, each lower code of the classification code and the technical field information corresponding to the lower code. A step of generating a commentary image including a second display portion that is displayed in association with each other, and a classification code assigned to the searched document, which corresponds to at least one of a higher code and a lower code, Emphasizing at least a part of the commentary image, thereby generating the tendency analysis image, and the aggregation processing step comprises: Set the compartment and to display the lower code in each compartment, thereby comprises the step of generating the trend analysis table, the display step,
When the upper code or the lower code on the tendency analysis table is selected, the portion of the tendency analysis image corresponding to the upper code or the lower code is scroll-displayed.

The present invention for solving the above-mentioned problems of the prior art is a program executed by a computer connected to a database that stores patent documents to which classification codes having a plurality of layers are added, and the program is input. A search step of searching for patent documents from the database according to the search conditions described above, and at least one obtained as a result of the search.
An image generation step of generating a tendency analysis image representing the appearance tendency of the classification code assigned to each document with respect to a document set consisting of two patent documents, wherein the classification code is an upper code representing a higher classification and a lower classification. And a lower code representing each of the upper codes of the classification code and the technical field information corresponding to the upper code in association with each other, and each of the lower codes of the classification code. A step of generating a commentary image including a second display unit that displays the related technical field information corresponding to the lower order code, and a higher order code of the classification code assigned to the searched document. And at least a part of the reference image corresponding to at least one of the subordinate codes, thereby generating the tendency analysis image. It is a symptom.

The present invention for solving the above-mentioned problems of the conventional example is a program executed by a computer connected to a database accumulating patent documents to which classification codes having a plurality of layers are added, and the program is input. A search step of searching for patent documents from the database according to the search condition, and at least 1 obtained as a result of the search.
And a totalization process step of performing a totalization process on a document set consisting of three patent documents using the classification code and creating a trend analysis table showing the result of the totalization process, wherein the classification code represents a higher-level classification. Including a high-order code and a low-order code representing a low-order classification, wherein the aggregation processing step sets a section for each of the high-order codes, displays the low-order code in each section, and thereby creates the trend analysis table. Is characterized by.

The present invention for solving the above-mentioned problems of the conventional example is a program executed by a computer connected to a database that stores patent documents to which classification codes having a plurality of layers are added, and the program is input. A search step of searching for patent documents from the database according to the search conditions described above, and at least one obtained as a result of the search.
An image generation step of generating a tendency analysis image showing the appearance tendency of the classification code assigned to each document for a document set consisting of one patent document, and at least one patent document obtained as a result of the search. At least one of the aggregation processing step of performing aggregation processing on the set using the classification code and creating a trend analysis table representing the result of the aggregation processing, the trend analysis image, and the trend analysis table is scroll-displayed. A display step, wherein the classification code includes an upper code indicating a higher classification and a lower code indicating a lower classification, and the image generating step includes a technique corresponding to each upper code of the classification code and the higher code. A first display unit for displaying the field information in association with each other, each sub code of the classification code, and technical field information corresponding to the sub code. A step of generating a manual image including a second display portion for displaying the page, and a classification code assigned to the searched document, which corresponds to at least one of a higher code and a lower code, Emphasizing at least a part of the manual image, thereby generating the tendency analysis image, and the aggregation processing step,
A step of setting a section for each of the upper codes, displaying the lower code in each section, and creating the trend analysis table thereby, the displaying step includes the upper code or the lower code on the trend analysis table. When a code is selected, the portion of the trend analysis image corresponding to the upper code or the lower code is scroll-displayed.

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described with reference to the drawings. A patent information search device 1 according to an embodiment of the present invention includes a control unit 11, a storage unit 12, a disk 13, a network control unit 14, and an input / output unit 1.
5, and is connected to a client 2 and a database 3 that stores patent documents via a network. There may be a plurality of clients 2.

The control unit 11 of the patent information retrieval device 1 realizes the retrieval means, the aggregation processing means, and the means for presenting the retrieval result of the present invention, and performs the processing in accordance with the program stored in the disk 13. This control unit 11
Also performs processing as a Web server, dynamically creates an electronic document in accordance with an instruction from the client 2 received by the network control unit 14, and
Reply to. Specifically, when a search instruction is received from the client 2 together with the search condition, a patent document satisfying the search condition is searched from the database 3, the search result is generated as a Web page, and the search result is generated by the client 2
To provide. Specific processing of the control unit 11 will be described in detail later.

The storage unit 12 operates as a work memory of the control unit 11. The disk 13 stores a program executed by the control unit 11. The network control unit 14 is connected to the network, and the control unit 1
In accordance with an instruction input from 1, data is transmitted to a predetermined destination via the network, and data arrived via the network is received and output to the control unit 11. The input / output unit 15 is a display, a keyboard, or the like, and receives the operation of a person in charge at the time of maintenance or the like, transmits the operation content to the control unit 11, and displays information on the display according to the instruction input from the control unit 11. To do.

The client 2 is an ordinary personal computer in which a Web browser is installed, and is connected to the patent information search device 1 via a network. The database 3 includes a storage device and a computer device for providing the contents of the storage device via a network. The storage device of the database 3 stores text data such as patent specifications, image data, and bibliographic information as patent documents. Here, the bibliographic information includes
It includes information on the applicant and the right holder, date / period information such as filing date, and classification code information such as international patent classification and FI, F-term (US class for US patent documents). The database 3 may also hold index data for full-text search.

Here, the processing of the control unit 11 will be described separately for the search processing, the image generation processing, and the totalization processing. First, the search process of the control unit 11 will be described. Control unit 1
1 provides the client 2 with a Web page including a text field for inputting search conditions and a “search” button associated with activation of a program for search processing. When the user of the client 2 inputs information such as a keyword, a document number, date information (application date etc.) as a search condition into a text field of this Web page and clicks a search button, the input information is changed to the network control unit. It is received at 14 and output to the control unit 11.

The control unit 11 searches the database 3 for patent documents including keywords, for example, based on the search conditions thus obtained. For this search method, a widely known full-text search technique can be used as it is. Then, a Web page including a list of patent documents obtained by the search is generated, and the Web page is provided to the client 2. Here, each patent document included in the list includes a predetermined number such as an application number and a publication number, a title (name of invention or invention), applicant, final disposition (finally what kind of administrative disposition Whether or not it has been submitted, that is, patent decision, decision of refusal, withdrawal pseudonym, etc.), etc., and preferably, at least a part of this information includes a text such as an application, a description, a drawing, an abstract of the patent document. Includes link settings to data and image data. With this link setting, when the user clicks, for example, the publication number of each patent document on the client 2, a Web page including the text data and image data of the publication that is the patent document is provided.

What is characteristic of the present embodiment is that the control unit 11 indicates on the classification code commentary screen what kind of classification code is assigned to the patent document searched in this way. A trend analysis image shown as a distribution of the assigned classification codes and a trend analysis table shown as a distribution on a table arranged in a predetermined order are generated. The image generation process and the totalization process for creating the trend analysis image and the trend analysis table will be described below.

[Aggregation Processing] First, the aggregation processing will be described. The control unit 11 creates a classification code (international patent classification,
2), the processing shown in FIG. 2 is started, and for the set of patent documents obtained by the search, the classification code attached to each patent document is acquired, The data is stored in the storage unit 12 (S1). Next, the storage unit 12
An upper part (upper code) is extracted from the classification code from the contents of (S2). Then, the appearance frequency of the higher order code is calculated and stored in the storage unit 12 (S3), and the higher order code obtained in the process S2 is sorted in the order of appearance frequency to create a higher order code list (S4). Further, the control unit 11 calculates the appearance frequency of each classification code and stores it in the storage unit 12 (S5). Then, the control unit 11 sets the display section corresponding to each upper code in this order based on the list of the upper codes sorted in the appearance frequency (S6). In HTML (HyperText Markup Language), this setting is performed by outputting tags that generate each column of the table. Although no display is actually performed at this stage, the state in which the display section is set is virtually as shown in FIG.
The state becomes as shown in (a).

Next, the control section 11 sorts the classification codes belonging to the higher order code corresponding to the display section in each display section in the order of (the lower order code of) the classification code (that is, sorting of the predetermined classification codes). An ordered list is generated, and a trend analysis table is generated (S7). This process S
In step 7, for example, a tag for generating another table is described in each column of the table generated in step S6. Then, the control unit 11 provides the table thus created as a Web page to the client 2 (S8) and completes the processing. This web page is shown in FIG.
It includes the trend analysis table shown in (b). Note that, as shown in FIG. 3B, it is preferable that each display section represents the value of the appearance frequency of the higher-order code corresponding to the display section (A), and the list of each classification code indicates that classification code. It represents the value of the appearance frequency (B).

The user of the client 2 can know the tendency of the classification codes attached to the documents retrieved by this display, and since the higher-order code portions are sorted in the order of appearance frequency, the most probable codes are displayed. Will be able to recognize quickly.

[Image Generation Processing] Next, the image generation processing of the control unit 11 will be described. The control unit 11 receives an instruction from the client 2 to display the manual of the classification code and generates a screen for displaying the manual. Specifically, as shown in FIG. 4, this commentary screen has a first column (P) in which a higher code is associated with a description representing the technical field of each higher code, and a lower code (or a classification code itself). And a second column (Q) in which each subordinate code or description representing the technical field of each classification code is associated. Since each column is large as a whole, it is preferable to display a part of the column as shown in FIG. 4 and provide a vertically movable interface for display, that is, so-called scroll display.
This scroll display is widely known as a word processor screen display or the like if the computer uses a so-called graphical user interface.

For the set of patent documents obtained by the search, the control unit 11 displays the first document on the manual screen for each document.
In the column, the color of the portion corresponding to the higher-order code of the classification code attached to the document is inverted and highlighted. Further, the control unit 11 inverts and highlights the color of the portion corresponding to the classification code (or lower code) attached to the document in the second column. In this way, the tendency of the classification code of the documents included in the set of patent documents obtained by the search is mapped on the manual, and the tendency analysis image is generated.

The control unit 11 posts the trend analysis image thus generated on the Web page and provides it to the client 2. The image here is so-called JPEG (Join
t Picture Experts Group) This refers to a scroll display screen or the like generated by Java (registered trademark) or the like, unlike image data.

Here, the control unit 11 associates each upper code and each classification code of this Web page with a "designate" button (R) for setting the code as a search condition. The user can include the corresponding code in the search condition by clicking the "specify" button.

Here, since the manuals are usually arranged in the predetermined order of the classification codes, as described above, the classification codes adjacent to each other in this manual correspond to the technologies similar to each other. Become. This provides the user with an opportunity to find a more appropriate classification code by referring to other classification codes in the vicinity of the highlighted classification code, and assists the search using the classification code more appropriately.

When the trend analysis table is generated in advance, the storage unit 12 stores the calculation result of the appearance frequency of the upper code and the classification code. The highlighted portion may be generated. In this case, a more appropriate classification code can be obtained by changing the highlighting mode according to the information on the appearance frequency calculated when creating the trend analysis table, for example, by changing the color density according to the appearance frequency. Support is provided to find them.

[Cooperation between trend analysis table and trend analysis image] The control unit 11 further adds a character string representing each classification code and upper code shown in the trend analysis table to the classification code on the trend analysis image. Set a link to the corresponding location. On the contrary, if the character strings representing the classification code and the higher order code on the tendency analysis image are on the tendency analysis table, a link to the position of the character string on the tendency analysis table is set to the character string. . This setting can be realized, for example, by setting an anchor tag if the table is created in HTML, and a class for scroll display if the scroll display of the trend analysis image is realized by Java (registered trademark). On the other hand, a method of moving the linked line to the display center can be implemented.

[Reuse for Search Condition] Further, the upper code and the classification code thus obtained can be used as they are as the search condition.
That is, the control unit 11 controls the W shown in FIG.
When one or more codes are designated from the higher-level code and the classification code shown on the eb page, the patent document with the code is searched from the database 3, and the list of the patent documents as the search result is included. A web page is created and the web page is provided to the client 2.

[Document Selection] Here, regarding the list of documents obtained as a result of the search by the keyword,
Although it has been explained that the totaling process is performed on all the documents included in the list, for example, as a result of the search by the keyword, a document clearly different from the target field of the search may be searched. In such cases, it is not preferable to target all documents. For example,
When using a keyword like "charged" while investigating a technology related to "electrophotography using a charged image,"
Documents such as "antistatic substance" that are clearly unrelated may be found, and in this case, it is preferable to target only those relating to "electrophotography" in the aggregation process.

When there are a large number of documents obtained as a result of the search, it is not preferable from the viewpoint of processing efficiency and the result display that each of them is the target of the aggregation processing.

Therefore, in the present embodiment, prior to the aggregation processing, the user is made to select a document to be the object of the aggregation processing, and the aggregation processing is performed on the selected document. That is, a toggle switch (check box) serving as a “bookmark” indicating whether or not each document is selected as a target of aggregation processing is provided while providing a list of search results by keywords or the like as shown in FIG. -Provide interface etc. (C). The user clicks this toggle switch to decide whether or not to select it, and gives an instruction to start the aggregation process.

The control unit 11 receives information indicating which document has been selected and information about which classification code the totalization process is to be performed, and receives an instruction to start the totalization process, and then, FIG.
The processing shown in is performed. In this case, in step S1, the designated classification code attached to the selected document is acquired.

[Creation of Correspondence Table] Further, the control section 11
In addition to the list shown in FIG. 3B, in order to show which document the classification code is assigned to, a list of documents is arranged in a predetermined axis direction, and the documents are sorted in the axis direction orthogonal to the list by the above-mentioned aggregation processing. It is also preferable to generate a table in which the classification codes are arranged in this order, and display a mark for each document at a position corresponding to the classification code (and upper code) given to the document (FIG. 6). Here, the table also includes a list of codes assigned to each document (D).

[Operation] As described above, according to the patent information retrieving apparatus of the present embodiment, a user who intends to retrieve using a code accesses the patent information retrieving apparatus 1 using the client 2 and sets the retrieval condition. input. The patent information search device 1 obtains a set of patent documents according to the search condition obtained from the client 2 via the network and provides the set of documents to the client 2 as a Web page. Here, the search condition may be a keyword or may directly specify the document number.

The user selects the target of the aggregation processing from this set, and transmits the content of the selection to the patent information search device 1. Also, the user selects which classification code to use, and transmits the selection to the patent information search device 1. Here, it is assumed that the F-term is selected as the classification code. The patent information retrieval device 1 acquires the F-term assigned to the selected patent document, and extracts the theme code, which is a higher-level code, from this. Then, for each theme code, the frequency of appearance in the selected document (first appearance frequency) is calculated, and the frequency of appearance in the selected document (second appearance frequency) is calculated for each F term.

Then, the higher-order codes are rearranged in the order of the first appearance frequency to set a section, and the F-terms belonging to the higher-order code are rearranged in each section in the order of the predetermined F-terms.
A tendency analysis table (X) corresponding to each second appearance frequency is generated. Then, a correspondence table (Y) indicating which document each F term is assigned is generated and displayed as shown in FIG. 6 in accordance with the display of the code.

The user refers to this tendency analysis table in order from the top, and the theme code with the highest appearance frequency
By referring to the appearance tendency of the F term, the F term of the technical field to be searched is selected, and the search of patent documents is further continued by using this F term as a search condition.

If the search results are not satisfactory,
Alternatively, when the classification code cannot be sufficiently specified in the trend analysis table shown in FIG. 6, the instruction manual is referred to, and an instruction to display the appearance tendency of the classification code is displayed on the instruction manual. That is, the generation of the trend analysis image is instructed. Then, the patent information retrieving apparatus 1 highlights the set of patent documents as a result of the retrieval, on the manual image, in the part corresponding to the higher-order code or the classification code given to the documents included in the set. The generated trend analysis image
It is included in the page and delivered to the client 2. The client 2 refers to the classification tendency mapped on the manual image, and selects a classification code serving as a search condition from above on the manual.

The effect of selecting from the manual in this way will be explained, for example, in the international patent classification for the sake of simplicity. G06F 17/14 (Fourier, Walsh or similar domain transformation) and G06F 17/16 ( When a large number of documents to which (matrix or vector calculation) is added are searched, on the trend analysis table,
These classifications are presented first, but G06F 17/1
5 (calculation of correlation function) is not presented unless the document to which it is attached is searched. However, if it is on the trend analysis image that is mapped on the manual, 17/14
The 17/15 sandwiched between 17 and 16/16 is easily found to assist in the selection of the appropriate classification code in an interrelated order.

[0050]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, an example will be described in which the patent information search device of the present invention is implemented by using an alpha server (trademark) manufactured by Compaq (trade name) and a search is performed using this.
As shown in FIG. 7, in the storage of the server 101, Japanese patent publications after 1993, publications of publications,
About 8 million patent documents, including patent gazettes and US patent gazettes after 1980, have been accumulated, and international patent classifications, FI, and F terms have been added to Japanese documents,
The US class has been assigned to US literature. Also,
This server is running on Tru64 UNIX (registered trademark).

Further, this server is connected to a local area network (LAN) via a hub and passes through a router connecting to the LAN of each business office so that the client 102 on the LAN of each business office can It is accessible.

The user of each client will execute the search in the following flow. That is, first, a subject to be searched is determined, a preliminary search is performed using a keyword or the like, and a bookmark is used to form a set of exemplary documents. Then, the tendency of assigning a classification code to each document in the set of these exemplary documents is analyzed, and this is examined using a trend analysis table and a trend analysis image. Then, a classification code that matches the original search subject is found, and this is included in the search conditions and a search is performed again (main search). Then, with respect to the set of documents obtained as a result of the main search, the “bookmark” is further utilized to select the document of interest, and this is used for the subsequent processing.

On the other hand, the obtained trend analysis table and the trend analysis image itself are also used for classification of in-house developed technology, and a so-called patent map (the status of other companies' applications and the status of their own applications is used in the technical fields and products). It can be useful for the generation of documents described by analyzing each element).

Specifically, "a photoconductor base material for reducing a charging sound (discharge sound) generated at the time of charging of a contact charging system which greatly contributes to downsizing and low ozone of electrophotography.
Alternatively, a technique relating to the photoconductor structure, but excluding the correspondence on the charging member side, is the subject of the investigation, and a case where the search is executed using the server will be described.

In the patent document search apparatus of this embodiment, the initial keyword search is a preliminary one, so
There may be omissions in documents searched by keywords corresponding to the survey subject. Therefore, a full-text search is executed for the full-text text under the AND (logical product) condition of the following keywords. That is, regarding the patent publications issued by the Japan Patent Office after 1993, "contact charging" and "photoreceptor"
The three keywords "noise" are searched by AND condition.

Therefore, when the client 102 instructs the server 101 to perform a keyword search, the server 101 provides the client 102 with a Web page as shown in FIG. Enter the keyword (A) and instruct the search.

The server 101 searches the documents in the storage under this condition, lists the patent documents obtained by the search as shown in FIG. 9 (B), and among the documents listed according to the instruction. The client 102 is provided with a Web page in which the entire text of the specified document and a drawing are displayed in a multi-window. Here, a check box for setting a “bookmark” is displayed in the column (B) in which the documents of FIG. 9 are listed, and by clicking this check box, the user of the client 102 gives a “bookmark”. (Corresponding to FIG. 5 of the embodiment). Then, for each document selected and listed as an exemplary document, the classification code type (for example, F term) to be analyzed is specified, and an instruction to analyze the tendency to assign the classification code is issued.

The server 101 creates a tendency analysis table showing the giving tendency and presents it as a Web page.
This presentation result is shown in FIG. 3B described as the embodiment.
Corresponding to the screen example of FIG. 10, and specifically, as shown in FIG. In this example, the display is made so as to show even the corresponding documents.

That is, in FIG. 10, the uppermost code is shown in the leftmost column, and the uppermost code is arranged from top to bottom in the order of appearance frequency. Also, in the next column to the right and the second column from the left, the lower codes (entire classification code) grouped by the partition set by each upper code are displayed, and the inside of this partition (within the grouping) is sorted by the classification code. They are arranged in order (default order). Further, the two columns on the right side show the values of the appearance frequencies of the upper code and the classification code, respectively.

The user of the client 102 may directly select the classification code used for the search from this screen. However, this may not be sufficient, since only those given to the cited documents are shown here. Therefore, the server 101 is further instructed to generate the trend analysis image. Then, the server 101
The tendency of the classification code assigned to the documents in the search results is
In order to visually represent the classification code on the manual, the classification code given to the document as the search result is referred to and the corresponding classification code portion on the manual image is highlighted.
Provide the eb page (Fig. 11).

In FIG. 11, (a) "2H00
3: Electrostatic charge ", (b)" 2H071: General electrophotographic device and case, element ", (c)" 2H035: Electrostatic charge / photosensitive member shape in electrophotography "are generated. . Looking at each trend analysis image, (a) "2H003:
In "electrophotographic charging", the F term associated with the contact charging roll is continuously displayed, so that the intended one can be easily found. For example, if you want to include "corona charging" in the survey range even with the same charging method, the F term that is not highlighted here (that is, there is nothing given in the example literature, and it can be selected in the trend analysis table). What did not exist) can be selected and included in the search conditions. Similarly, in (b) “2H071: General electrophotographic apparatus, housing, and element”, F terms related to the image carrier are collectively highlighted, and (c) “2H035: Static elimination in electrophotography. Regarding the “photoreceptor shape”, the F terms related to the support are collectively highlighted.

Then, from this manual (trend analysis image), by clicking the "specify" button, the F-term as the search condition is selected. Here, "2H035 / CA07" and "2H035 / CA07"
035 / CB02 ”,“ 2H035 / CB03 ”, and“ 2H035 / CB04 ”are set as the first condition specified by OR (logical sum) condition, and“ 2H071 / DA06 ”and“ 2H071 / DA15 ”are set. OR (logical sum)
The second condition specified by the condition is set. These first and second conditions are connected by AND (logical product), and form a search condition as a whole.

When a search is executed under this condition, about 43
00 patent documents were searched for "Noise, Vibration,
By performing a narrowed search with keywords that combine "vibration prevention, silent noise, noise insulation, and electrostatic noise" with OR, approximately 400 patent documents are obtained as search results (FIG. 12).

The search results are provided as a Web page as shown in FIG. 13 corresponding to FIG. 9. Here, the column (B) showing a list of documents is enumerated as an example at the time of preliminary search. The "bookmark" information given by is maintained. That is, the “bookmark” previously given is also displayed in this field.

With respect to the document to which the “bookmark” is added in this way, the server 101 generates a Web page (FIG. 14) listing a so-called abstract including an abstract according to an instruction. In addition, a screen that displays more detailed information from this abstract to a part that represents a specific document such as a title (see FIG. 1).
The link to 5) is preferably set. The screen of FIG. 15 is provided with a column in which the user who has searched makes a comment on the document concerned, and the comment column may be shared between groups of predetermined users, or the storage of the server 101. May be recorded in.

Further, the server 101 provides a field for each technical field in the direction of a predetermined axis based on the classification code, and a field according to the application (for example, year of application) in a direction orthogonal to the predetermined axis for each technology. It is preferable to further include means for generating a patent map image in which the field and the application trend for each application year are commented (FIG. 16).

As a result, from the results of the investigation, for example, what means / material /
You can get a panoramic view of the companies that have acquired the rights in the composition, including the trends of your company and other companies, and you can use it for patent monitoring of other companies (so-called SDI) as it is.
The results can be combined and used in patent strategies.

Further, in order to create an in-house classification (in-house technical system) and extract points of the invention, the in-house development company itself participates in this comment creation and the like, thereby obtaining a new guideline for technological development through patent research. It becomes possible to use it widely.

Next, a case where a US patent search is performed using the server 101 of this embodiment will be described. In this case also, the search is performed in the same flow. That is, first, the keyword is preliminarily searched (FIG. 17). Here, "Photoreceptor", "Silencer", and "contac" are selected from about 2.23 million US patent documents from 1980 to about July 2001.
The logical product of "t", "charg%", and "roll%" was used as the search condition. In addition, since there is a possibility that the ending may change in a search in English, a character that matches any 0 or more characters using a so-called regular expression, such as "roll%", is used. % "Etc. are also included in the specification. As a result, not only "roll" but also ending change states such as "rolling" can be included in the search.

As a result of this search, one patent document is searched, and it is confirmed by displaying the document whether or not it matches the target technical field (see FIG. 1).
8) If correct, add a "bookmark" to it.

When the trend analysis table is generated,
The server 101 generates a trend analysis table as shown in FIG. 19 and provides it as a Web page. Therefore, these classification codes "399/91", "399/109", "399/159"
Regarding the above, the server 101 is instructed to provide the explanation image of the classification code, and the explanation is referred to (FIG. 20).
The logical product of "1" and "399/159" is set as the classification code of the search condition.

As a result, since four documents are searched,
Furthermore, the contents of these four cases are referred to confirm that they are included in the survey target, and "bookmark" is given to all four cases to further generate the trend analysis table. That is, in this example, the search result using the classification code is also used as the preliminary search. However, since a new classification code does not appear even if the trend analysis table is referenced, "399/91" and "399/91"
The search condition connected by the logical sum of "159" is newly set and the search is performed (FIG. 21). Then, 374 documents are obtained. In this way, a search that effectively uses the classification code is performed while easily changing the search condition.

[0073]

According to the present invention, the tendencies of the codes assigned to the respective patent documents can be visually displayed on the trend analysis table or the trend analysis image for the designated patent documents. Since the classification code found as a result of the analysis can be specified as the search condition, it is possible to support the search effectively using the classification code.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration and a connection state of a patent information search device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing an example of a counting process in the control unit 11.

FIG. 3 is an explanatory diagram showing an example of a display state of a trend analysis table obtained as a result of the aggregation processing.

FIG. 4 is an explanatory diagram illustrating a display example of a trend analysis image.

FIG. 5 is an explanatory diagram illustrating an example of a display state of a search processing result.

FIG. 6 is an explanatory diagram showing another example of the display state of the trend analysis table obtained as a result of the aggregation processing.

FIG. 7 is a block diagram showing a configuration of a patent information search device according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 9 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 10 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 11 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 12 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 13 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 14 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 15 is an explanatory diagram showing a display example of a Web page provided from a patent information search device according to an embodiment of the present invention.

FIG. 16 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 17 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 18 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 19 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 20 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

FIG. 21 is an explanatory diagram illustrating a display example of a Web page provided by the patent information search device according to the embodiment of the present invention.

[Explanation of symbols]

1 Patent information retrieval device, 2, 102 clients, 3
Database, 11 control unit, 12 storage unit, 13 disk, 14 network control unit, 15 input / output unit, 1
01 server.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Hidetoshi Osafune             Fuji Zero, 6-20 Akasaka, Minato-ku, Tokyo             X Co., Ltd. F term (reference) 5B075 ND20 NK02 NK43 PQ02 PQ15                       PQ22 PQ40 PQ66

Claims (12)

[Claims] 1. A search means for searching for patent documents from a database accumulating patent documents to which classification codes having a plurality of layers are added, and at least one patent obtained as a result of the search. An image generation unit that generates a tendency analysis image that represents the appearance tendency of the classification code assigned to each document with respect to a document set made up of documents, and the classification code includes a high-order code indicating a high-order classification and a low-order classification. A lower display code, and the image generation means, a first display unit for displaying each higher code of the classification code and technical field information corresponding to the higher code in association with each other, and a lower code of the classification code, A means for generating a manual image including a second display unit that displays the technical field information corresponding to the lower order code in association with the second code; Regarding the classification code assigned to the literature,
A patent information retrieving apparatus comprising: a means for emphasizing at least a part of the manual image corresponding to at least one of the upper code and the lower code, thereby generating the tendency analysis image. 2. A search means for searching a patent document according to an input search condition from a database accumulating patent documents to which classification codes having a plurality of layers are assigned, and at least one obtained as a result of the search. Aggregation processing means for performing aggregation processing on the document set consisting of patent documents using the classification code, and creating a trend analysis table showing the result of the aggregation processing, wherein the classification code represents the upper classification A means for creating a trend analysis table by including a higher-order code and a lower-order code representing a lower-order classification, wherein the aggregation processing means sets a section for each of the higher-order codes, displays the lower-order code within each section A patent information search device comprising: 3. A search means for searching a patent document according to an input search condition from a database accumulating patent documents to which a classification code having a plurality of layers is assigned, and at least one patent obtained as a result of the search. An image generation unit that generates a tendency analysis image that represents the appearance tendency of a classification code assigned to each document with respect to a document set that includes documents, and a document set that includes at least one patent document obtained as a result of the search. On the other hand, a totalization processing unit that performs totalization processing using the classification code and creates a trend analysis table representing the totalization processing result, a display unit that scrolls and displays at least one of the trend analysis image and the trend analysis table. And the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, The forming means includes a first display unit for displaying each upper code of the classification code and the technical field information corresponding to the higher code in association with each other, each lower code of the classification code and the technical field information corresponding to the lower code. A means for generating a commentary image including a second display unit that displays and in association with each other, and a classification code assigned to the searched document,
A means for emphasizing at least a part of the manual image corresponding to at least one of the upper code and the lower code, thereby generating the tendency analysis image, wherein the totalization processing means includes the upper code A section for setting each section and displaying the lower order code in each section, thereby creating the trend analysis table, wherein the display section selects the upper order code or the lower order code on the tendency analysis table. Then, the patent information retrieving apparatus is characterized in that a portion of the trend analysis image corresponding to the upper code or the lower code is scroll-displayed. 4. The patent information search device according to claim 2 or 3, wherein the aggregation processing unit calculates a first appearance frequency for each higher-order code found from the patent document subset, and A means for calculating a second appearance frequency based on a classification code found in the patent document subset, the display order of the sections is adjusted based on the first appearance frequency, and based on the second appearance frequency. A patent information retrieval device characterized in that the display order of lower order codes according to each classification code is adjusted in each section. 5. The patent information retrieval apparatus according to claim 1, further comprising a unit for designating a classification code as a retrieval condition on the tendency analysis image or the manual image, and the designation. A patent information retrieving apparatus, wherein the retrieved classification code is provided for retrieval processing as at least a part of retrieval conditions. 6. The patent information search device according to claim 1, wherein the search unit performs a preliminary search based on an input search condition, and a result of the preliminary search. A means for outputting at least one selected patent document obtained as a result of the search by an instruction to select a part of the obtained document, and at least one output as the result of the search. A patent information retrieval device characterized in that patent documents are subjected to predetermined processing as the document set. 7. A search step of searching for patent documents according to an input search condition from a database accumulating documents to which classification codes having a plurality of layers are added, and at least one patent document obtained as a result of the search. An image generation step of generating a tendency analysis image representing the appearance tendency of the classification code assigned to each document with respect to the document set consisting of, wherein the classification code represents an upper code representing a higher classification and a lower classification. A lower display code, the image generation step, each upper code of the classification code and a first display unit for displaying the technical field information corresponding to the higher code in association with each other, the lower code of the classification code and A step of generating a manual image including a second display section that displays the technical field information corresponding to the lower order code in association with each other; Regarding the classification code assigned to
And a step of emphasizing at least a part of the commentary image corresponding to at least one of the higher order code and the lower order code, thereby generating the tendency analysis image. 8. A search step of searching for patent documents according to input search conditions from a database accumulating patent documents to which classification codes having a plurality of layers have been added, and at least one obtained as a result of the search. A totalization processing step of performing a totalization process on a document set consisting of patent documents using the classification code, and creating a trend analysis table representing the result of the totalization process, wherein the classification code represents a higher classification Including a high-order code and a low-order code representing a low-order classification, wherein the aggregation processing step sets a section for each of the high-order codes, displays the low-order code in each section, and thereby creates the trend analysis table. Patent information retrieval method characterized by: 9. A search step of searching for patent documents according to an input search condition from a database accumulating patent documents to which classification codes having a plurality of layers are assigned, and at least one patent obtained as a result of the search. An image generation step of generating a tendency analysis image showing the appearance tendency of the classification code assigned to each document for a document set consisting of documents, and a document set consisting of at least one patent document obtained as a result of the search. On the other hand, a totalization process that performs a totalization process using the classification code and creates a trend analysis table that represents the result of the totalization process, and a display process that scrolls and displays at least one of the trend analysis image and the trend analysis table. And the classification code includes an upper code representing a higher classification and a lower code representing a lower classification, The forming step includes a first display section for displaying each higher code of the classification code and the technical field information corresponding to the higher code in association with each other, each lower code of the classification code, and technical field information corresponding to the lower code. A step of generating a commentary image including a second display unit that displays and in association with each other, and a classification code assigned to the searched document,
A step of emphasizing at least a part of the commentary image corresponding to at least one of the higher order code and the lower order code, thereby generating the tendency analysis image; Setting a section for each of the sections, displaying the lower order code in each section, thereby creating the trend analysis table, wherein the displaying step selects a higher order code or a lower order code on the trend analysis table. Then, the patent information retrieval method, wherein a portion of the tendency analysis image corresponding to the upper code or the lower code is scroll-displayed. 10. A search step of searching a patent document from the database according to an input search condition in a computer connected to a database that stores the patent document to which a classification code having a plurality of layers is added, and a result of the search. The image generation step of generating a tendency analysis image representing the appearance tendency of the classification code assigned to each document for the document set consisting of at least one patent document obtained as The code includes an upper code indicating a higher classification and a lower code indicating a lower classification, and the image generating step displays each upper code of the classification code and technical field information corresponding to the higher code in association with each other. 1 display unit, and displays each lower code of the classification code and the technical field information corresponding to the lower code in association with each other. Generating a Reference image comprising a second display unit, the assigned classification code in the retrieved documents,
And a step of emphasizing at least a part of the manual image corresponding to at least one of the upper code and the lower code, thereby generating the tendency analysis image. 11. A search step of searching a patent document from the database according to an input search condition in a computer connected to a database that stores a patent document to which a classification code having a plurality of layers is added, and a result of the search. A totalization process for performing a totalization process on the document set consisting of at least one patent document obtained as above using the classification code, and creating a trend analysis table representing the totalization process result. Wherein the classification code includes an upper code indicating a higher classification and a lower code indicating a lower classification, and the aggregation processing step sets a section for each of the upper codes and displays the lower code in each section. And a program for creating the tendency analysis table. 12. A search step of searching a patent document from the database according to an input search condition in a computer connected to a database that stores a patent document to which a classification code having a plurality of layers is assigned, and a result of the search. An image generation step of generating a tendency analysis image representing the appearance tendency of the classification code assigned to each document for the document set consisting of at least one patent document obtained as above, and at least 1 obtained as a result of the search. For a document set consisting of two patent documents, a totalization process of performing a totalization process using the classification code and creating a trend analysis table representing the result of the totalization process, the trend analysis image, and a trend analysis table A program for executing a display step of scrolling and displaying at least one of the above, wherein the classification code is A first display unit that includes a high-order code indicating a classification and a low-order code indicating a low-order classification, wherein the image generation step displays each high-order code of the classification code and technical field information corresponding to the high-order code in association with each other; A step of generating a manual image including a second display section that displays each subordinate code of the classification code and the technical field information corresponding to the subordinate code in association with each other; About the classification code,
A step of emphasizing at least a part of the commentary image corresponding to at least one of the higher order code and the lower order code, thereby generating the tendency analysis image; Setting a section for each of the sections, displaying the lower order code in each section, thereby creating the trend analysis table, wherein the displaying step selects a higher order code or a lower order code on the trend analysis table. Then, a program for scrolling and displaying a portion of the trend analysis image corresponding to the upper code or the lower code.