Information retrieval device, information retrieval method and information retrieval program

Abstract

To provide an information retrieval device which displays many pieces of data such as patents, essays, and merchandise as an integrated map and can perform retrieval while being interactively operated.SOLUTION: An information retrieval device 1 in the present invention visualizes retrieval data consisting of retrieval object information and key information regarding the retrieval object information on a map to retrieve the retrieval object information, and comprises: an integrated map generation unit 20 which maps a model estimated on the basis of a plurality of different multidimensional viewpoints in a lower dimension to generate an integrated map about the retrieval data; and an integrated map control unit 25 which performs control to display a global map for every viewpoint when the integrated map is viewed from the respective viewpoints of the retrieval object information and the key information, and to display, and when an arbitrary element on the displayed global map is selected, performs control to display a local map when the integrated map is viewed from viewpoints of the retrieval object information or the key information regarding the selected element and an element located near the selected element.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an apparatus, method and program for searching for information of interest to a user from a large number of searchable information such as patents, papers, merchandise, music and books present in databases or on a network. [Background technology]

[0002] Conventionally, when searching for target information from a large amount of publicly available information such as patents, there is a need to find all of the information of interest, rather than just one piece of information.

[0003] However, for example, when searching a field for the first time using a patent search, it is often the case that users do not know what patents have already been published or what keywords to use in their search. Also, when venturing into a new research field and searching for papers, they often do not know the research trends or keywords in that field. Furthermore, when traveling to a certain region for the first time, users often do not know where things are located, what famous places there are, or what delicious food there is, and when searching for such information, they often do not know what keywords to use to accurately find information about their destination.

[0004] Patent Document 1, by the inventor of the present applicant, discloses an integrated map creation device that, for each piece of information viewed from multiple different viewpoints, estimates a model of the information as viewed from another viewpoint, and generates and displays the estimated model as an integrated map. This integrated map displays information from various viewpoints in association with information viewed from other viewpoints, allowing analysis to be performed while interactively operating according to the user's intentions, and can be used to search for desired information from a large number of search target information such as patents. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 5750804 Summary of the Invention [Problem to be solved by the invention]

[0006] The present invention provides an information search device that displays search target information such as numerous patents, papers, and other literature, products, music, and books as an integrated map that associates information from one perspective with information viewed from another perspective, and allows users to perform searches while operating interactively in line with their intentions. [Means for solving the problem]

[0007] The information retrieval device of the present invention is an information retrieval device that visualizes search data consisting of search target information and key information related to the search target information on a map and searches for the search target information, and is equipped with an integrated map generation unit that generates an integrated map by mapping a model estimated based on a plurality of different multidimensional perspectives for the search data to a lower dimension, and an integrated map control unit that controls the integrated map to display a global map for each perspective when viewed from each of the perspectives of the search target information and the key information, and when any element on the displayed global map is selected, controls to display a local map when viewed from the perspective of the search target information or key information related to the selected element and elements located in its vicinity.

[0008] The information retrieval method of the present invention is an information retrieval method that visualizes search data consisting of search target information and key information related to the search target information on a map and searches for the search target information, and includes an integrated map generation step of mapping a model estimated based on a plurality of different multidimensional viewpoints for the search data to a lower dimension to generate an integrated map, and an integrated map control step of controlling the integrated map to display a global map for each viewpoint when viewed from each viewpoint of the search target information and the key information, and when any element on the displayed global map is selected, controlling to display a local map when viewed from the viewpoint of the search target information or key information related to the selected element and elements located in its vicinity. [Effects of the Invention]

[0009] According to the present invention, search target information such as a large number of patents can be displayed on a global map by associating information from one perspective with information viewed from another perspective, and any element on the global map can be selected to display a local map, allowing users to find search target information or key information of interest through a series of operations. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing a configuration of an information search device according to a first embodiment. [Figure 2] FIG. 1 shows patent data related to autonomous driving in the first embodiment. [Figure 3] FIG. 10 is a diagram showing an example of a display of an integrated map of patent data in the first embodiment. [Figure 4] Flowchart showing the operation of the information search device in the first embodiment [Figure 5] FIG. 10 is a diagram showing an example of a display of a global map of patent data in the first embodiment. [Figure 6] FIG. 10 is a diagram showing a display example when an element is selected in the global patent map according to the first embodiment. [Figure 7] FIG. 10 is a diagram showing a display example when an element is selected from the local keyword map according to the first embodiment. [Figure 8] FIG. 10 is a diagram showing a display example when an element is selected from the global keyword map according to the first embodiment. [Figure 9] FIG. 10 is a diagram showing a display example when an element is selected in the local patent map according to the first embodiment. [Figure 10] FIG. 10 is a diagram showing a display example when another element is selected in the global patent map according to the first embodiment. [Figure 11] FIG. 10 is a diagram showing a display example when another element is selected and registered as a favorite on the global patent map in the first embodiment. [Figure 12] FIG. 10 is a diagram showing a display example when another element is selected and registered as a favorite in the global keyword map according to the first embodiment. [Figure 13] FIG. 10 shows an example of a My Map display in the first embodiment. [Figure 14] FIG. 10 is a diagram showing the change in the display example of the local keyword map when elements are successively selected in the global patent map in the first embodiment. [Figure 15] Flowchart showing the operation of the information search device in the second embodiment [Figure 16] FIG. 10 is a diagram showing an example of a display of an integrated map of patent data in the second embodiment. [Figure 17] FIG. 10 is a diagram showing a display example when a certain range is selected in the global patent map in the second embodiment. [Figure 18] FIG. 11 is a diagram showing a display example when a certain range is selected in the global keyword map according to the second embodiment. [Figure 19] FIG. 10 is a diagram showing a display example when an element is selected in the local patent map according to the second embodiment. [Figure 20] FIG. 11 is a diagram showing a display example when an element is selected from a local keyword map according to the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present invention will be described, but the present invention is not limited to the following embodiments.

[0012] (Embodiment 1) [1. Configuration] 1 is a block diagram showing the configuration of an information retrieval device 1 according to this embodiment. The information retrieval device 1 includes an integrated map unit 2 that generates and controls an integrated map, an aggregated data storage unit 3 that stores multidimensional array nonlinear data related to search target information such as a large number of patents searched by the information retrieval device 1, an instruction unit 4 that gives instructions to the integrated map unit 2 based on information input by a user, and a display unit 5 that displays information such as maps.

[0013] The integrated map unit 2 includes an integrated map generating unit 20 that generates an integrated map, and an integrated map control unit 25 that controls the integrated map generating unit 20 and displays the integrated map on the display unit 5.

[0014] The integrated map generation unit 20 includes an initialization processing unit 21 that provisionally sets the placement positions of elements of individual maps based on the placement configuration of each element for each viewpoint obtained by learning the data in the aggregated data storage unit 3; an individual map information generation unit 22 that reads the placement positions of the individual maps and the data stored in the aggregated data storage unit 3 to generate individual maps and outputs the generated individual maps to a placement position calculation unit 23; a placement position calculation unit 23 that calculates the positions of the individual maps in the integrated map; and an integrated map information generation unit 24 that reads the placement configuration for each viewpoint calculated by the placement position calculation unit 23 and the data stored in the aggregated data storage unit 3 to generate an integrated map.

[0015] The information retrieval device 1 can be realized by, for example, a computer configured with a memory and a processor. The integrated map unit 2, the integrated map generation unit 20, the initialization processing unit 21, the individual map information generation unit 22, the layout position calculation unit 23, the integrated map information generation unit 24, and the integrated map control unit 25 are realized by software in which the memory and the processor operate in cooperation. Note that the integrated map unit 2, the integrated map generation unit 20, the initialization processing unit 21, the individual map information generation unit 22, the layout position calculation unit 23, the integrated map information generation unit 24, and the integrated map control unit 25 may be realized not only by software but also partially by hardware.

[0016] The instruction unit 4 is composed of a mouse and keyboard that accept input operations from the user, and software that processes the accepted input information. The display unit 5 is typically a display, but may also output to a printer. The aggregated data storage unit 3 is composed of a storage medium that stores data, such as an HDD or SSD, and a relational database that operates on the storage medium. Note that, although this embodiment describes a form in which the aggregated data storage unit 3 exists inside the information retrieval device 1, the information retrieval device 1 may also be equipped with a communication unit and acquire data present on an external server or the like via the communication unit.

[0017] [2-1. Basic model operation] First, we will explain the operation of the basic model of the information retrieval device 1, which estimates a model of information viewed from multiple different viewpoints as it would appear when viewed from another viewpoint, and generates and displays the estimated models as an integrated map. In explaining the operation of the basic model, we will use patent data that collects patents related to autonomous driving.

[0018] The aggregate data storage unit 3 stores patent data shown in Figure 2. As shown in Figure 2, the patent data includes information about the patent and keywords that describe the characteristics of the patent. The information about the patent includes the patent number for uniquely identifying the patent, the publication number that is the number when the patent was published, the filing year of the patent application, the country in which the patent was filed, etc. In Figure 2, a "1" is set if the patent shown in a row contains each keyword shown in a column, and a "0" is set if the keyword is not included. For example, the patent with patent number "JP41034" contains the keywords "distance measurement," "automatic braking," and "collision avoidance," but does not include "camera," "hands-free," or "automated driving."

[0019] The patent data shown in Figure 2 indicates "1" if the patent contains a keyword and "0" if it does not, but it may also be possible to count the number of times the keyword appears in the patent and include a number indicating that number as a parameter of the patent data. Furthermore, the method for extracting the keywords for each patent may be any existing keyword extraction method, such as extracting nouns that appear in abstracts. For the sake of convenience, Figure 2 shows only six patent data items, but the data volume is assumed to be tens of thousands of patent data items.

[0020] The integrated map generator 20 not only generates a map by mapping multidimensional nonlinear data to a lower dimension, but also generates a map by mapping a model estimated based on multiple different viewpoints and taking into account the relationships between those viewpoints into a lower dimension. That is, instead of generating a map based on different viewpoints (e.g., patents, keywords) from only the information classified at those viewpoints, the integrated map generator 20 generates a map in which the information classified at each viewpoint is consistently and appropriately associated. Therefore, when viewing a generated map from a certain viewpoint, it is possible to refer to information that takes into account the relationships with other viewpoints. Even with large-scale data in which multiple elements (e.g., patents "JP41034, JP51555,...") are intricately intertwined, it is possible to refer to information that takes into account each element and the relationships between each element.

[0021] In this embodiment, the integrated map generator 20 generates the integrated map by repeating learning using, for example, the EM algorithm. Note that the EM algorithm is just one method, and the integrated map may also be generated by repeating learning using a gradient method with an objective function.

[0022] The initialization processing unit 21 provisionally sets the layout positions of the elements of the individual maps based on the layout configuration of each element for each viewpoint obtained by learning the data in the total data storage unit 3.

[0023] The individual map information generation unit 22 generates an individual map by reading the placement position of the individual map calculated by the placement position calculation unit 23 (in the initial state, the placement position set by the initialization processing unit 21) and the data stored in the aggregated data storage unit 3. At this time, an individual map of patents for each keyword (for example, a map of patents for cameras) is generated from the placement positions of the patents obtained by the calculation in the above example and the patent data for each keyword stored in the aggregated data storage unit 3, for example.

[0024] The placement position calculation unit 23 calculates the position of the individual map in the integrated map. In the initial state, the placement position of the elements for each viewpoint is set by the initialization processing unit 21, so no processing is performed here. When the process moves from the initial state to the second or subsequent processing, the integrated map information generation unit 24 generates an integrated map, and the individual map information generation unit 22 generates an individual map, so the integrated map and the individual map are input, and the position of the individual map in the integrated map is calculated. At this time, for example, when calculating the position of an individual map of a keyword generated for each patent (e.g., a map of keywords for JP41034), the position of the patent is calculated by determining the position where the individual maps of the keyword are similar.

[0025] The integrated map information generation unit 24 generates an integrated map by reading the positions of the individual maps for each viewpoint calculated by the placement position calculation unit 23 (in the initial state, the placement positions set by the initialization processing unit 21) and all data stored in the aggregate data storage unit 3. That is, for elements identified based on all placement positions for each element at each viewpoint calculated by the placement position calculation unit 23 (for example, in the above example, the placement position identified by the placement position of JP41034 and the placement position of the camera (placement position of JP41034 × placement position of the camera)), actual data is input, and a single integrated map is generated in which all elements at all viewpoints have been learned.

[0026] The integrated map control unit 25 controls the display unit 5 to display the completed integrated map in accordance with instructions from the instruction unit 4, which are instructions from the user. The integrated map completed by the above processes has individual maps for each viewpoint, which are two-dimensional. Therefore, if there are two or more viewpoints, the integrated map becomes a multidimensional integrated map of four or more dimensions, which cannot be displayed as useful information. Therefore, it is necessary to represent the multidimensional integrated map in two dimensions. Furthermore, by fixing any parameters in accordance with instructions from the user, the necessary information can be output in an easy-to-read format. Even if the number of viewpoints increases, the basic algorithm remains the same as that described above, and it is possible to generate two or more viewpoints by taking into account the placement positions of all other viewpoints when generating the individual maps.

[0027] An example of an integrated map is shown below. Figure 3 simultaneously displays the integrated map as viewed from viewpoint 1 (patent map) and as viewed from viewpoint 2 (keyword map). Each map (each map viewed from a different viewpoint is considered a single-viewpoint map) is an integrated map modeled with each viewpoint in mind, so a single-viewpoint map from one viewpoint takes the other viewpoints into account. Furthermore, in each single-viewpoint map, elements with high similarity are placed close to each other. For example, in the patent map, "JP79123" is more similar to "JP70035" than to "JP78235," and in the keyword map, "camera" is more similar to "automatic braking" than to "eyes off."

[0028] Although the integrated map in this embodiment is realized by using the EM algorithm, other techniques such as SOM (Self-Organization Maps), GTM (Generative Topographic Map), topology-preserving mapping, manifold learning, or nonlinear subspace method can also be used.

[0029] [2-2. Search operation using global and local maps] Next, a search operation using the global map and the local map will be described with reference to the flowchart of FIG.

[0030] The integrated map generating unit 20 reads all patent data stored in the aggregated data storage unit 3 (step S401) and generates an integrated map related to the patent data (step S402). The operation of generating the integrated map is the same as the operation described in the basic model operation. The initialization processing unit 21, individual map information generating unit 22, layout position calculation unit 23, and integrated map information generating unit 24 that constitute the integrated map generating unit 20 work together to generate the integrated map by repeating learning using, for example, an EM algorithm.

[0031] In response to an instruction from the user via the instruction unit 4 to display the integrated map, the integrated map control unit 25 controls the display unit 5 to display the generated integrated map covering all patent data (step S403). The display screen of the display unit 5 is shown in Figure 5. The display screen is composed of four screens: a global patent map, a global keyword map, a local keyword map, and a local patent map. The global map displays the integrated map covering all patent data. The global patent map is the integrated map covering all patent data viewed from the perspective of patents, and the global keyword map is the map viewed from the perspective of keywords. The local keyword map is created by selecting any element from the global patent map and displaying keywords related to the selected element. The local patent map is created by selecting any element from the global keyword map and displaying patents related to the selected element.

[0032] Next, we will explain the operation when a user selects an arbitrary element on the global map. Figure 6 shows an example of the screen display when a user selects an element on the global patent map. When the user uses the instruction unit 4 to select patent "JP41034" on the global patent map displayed on the display unit 5 (step S404), keywords contained in "JP41034" and patents located nearby are displayed on the local keyword map (step S405).

[0033] On the global patent map, patents containing the same keywords are placed closer together. Specifically, the more similar a patent is to patent "JP41034" and contains the same keywords as those contained in patent "JP41034," the closer it is to patent "JP41034" on the global patent map. When a user selects patent "JP41034" on the global patent map, not only are the keywords contained in patent "JP41034" displayed, but also the keywords contained in patents located near patent "JP41034" that are highly similar in content.

[0034] Users can select the extent to which patents located near the selected patent should be included. For example, there are various methods, such as including patents that contain a preset percentage or more of the same keywords as patent "JP41034," or quantifying the similarity between patents and including patents whose numerical value is equal to or greater than a preset value. Users can check keywords related to the selected patent "JP41034" by referring to the local keyword map.

[0035] When the user refers to the local keyword map and finds a keyword of interest, the user selects the corresponding keyword on the local keyword map (step S409). FIG. 7 shows an example of a screen display when the user selects a keyword on the local keyword map. When a keyword is selected on the local keyword map, the corresponding keyword is explicitly displayed on the global keyword map (step S410). To explicitly display a keyword on the global keyword map, the integrated map control unit 25 changes the background color of the corresponding keyword. The method of explicitly displaying the keyword is not limited to this method, and other methods may be used, such as blinking the corresponding keyword or changing the character color of the keyword.

[0036] In this embodiment, a patent is selected on the global patent map, a local keyword map is displayed, and then when a keyword of interest is selected on the local keyword map, the corresponding keyword on the global keyword map is explicitly displayed. The operation of selecting a keyword of interest on the local keyword map does not have to be performed after updating the display of the local keyword map, but may be performed after updating the local patent map by performing an operation on the global keyword map. As long as the local keyword map is displayed and a keyword can be selected, the operation may be performed at any time.

[0037] The user selects a keyword that is explicitly displayed on the global keyword map. Figure 8 shows an example of the screen display when the user selects an element on the global keyword map. If the user is interested in the keyword "collision avoidance" and uses the instruction unit 4 to select the keyword "collision avoidance" on the global keyword map (step S411), patents that include "collision avoidance" and keywords located nearby will be displayed on the local patent map (step S412).

[0038] On the global keyword map, keywords contained in similar patents are placed closer together. Specifically, among patents containing the keyword "collision avoidance," the more of the same keyword that is contained in patents with high content similarity, the closer the keywords are placed to the keyword "collision avoidance" on the global keyword map. When a user selects the keyword "collision avoidance" on the global keyword map, not only patents containing the keyword "collision avoidance" but also patents containing keywords similar to the keyword "collision avoidance" are displayed.

[0039] Users can select the extent to which highly similar keywords should be included. For example, there are various methods, such as including keywords if a preset percentage or more of patents in a group of patents containing the keyword "collision avoidance" contain the same keyword, or quantifying the similarity between keywords and including keywords whose numerical value is equal to or greater than a preset value.

[0040] When a user refers to the local patent map and finds a patent of interest, the user selects the corresponding patent on the local patent map (step S416). Figure 9 shows an example of the screen display when a user selects a patent on the local patent map. When a patent is selected on the local patent map, the corresponding patent is explicitly displayed on the global patent map (step S417). To explicitly display a patent on the global patent map, the integrated map control unit 25 changes the background color of the corresponding patent. The method of explicitly displaying is not limited to this method, and other methods may be used, such as flashing the corresponding patent or changing the font color of the patent.

[0041] In this embodiment, a form has been described in which a patent is selected on the global keyword map, a local patent map is displayed, and then a keyword of interest is selected on the local patent map. However, the operation of selecting a patent of interest on the local patent map can be performed at any time as long as the local patent map is displayed and a patent can be selected.

[0042] The user selects an explicitly displayed patent on the global patent map. Figure 10 shows an example of the screen display when the user selects an explicitly displayed element on the global patent map. When the user selects patent "JP89222" (step S404), the local keyword map is updated and keywords contained in "JP89222" and patents located nearby are displayed (step S405).

[0043] In this way, users can find patents and keywords of interest by repeating the process of selecting any patent on the global patent map, checking keywords related to the selected patent on the local keyword map, selecting any keyword on the global keyword map, and checking patents related to the selected keyword on the local patent map.

[0044] In this embodiment, a configuration has been described in which, when a user selects a patent or keyword of interest on a local map, the same patent or keyword on the global map is explicitly displayed. When a user selects a keyword on the local keyword map, the same keyword may be selected on the global keyword map, and patents containing the keyword located nearby on the global keyword map may be displayed on the local patent map. Also, when a user selects a patent on the local patent map, the same patent may be selected on the global patent map, and keywords included in patents located nearby on the global patent map may be displayed on the local keyword map.

[0045] A user selects a patent on the global patent map, checks the keywords related to the selected patent on the local keyword map, and, if the selected patent is of interest, registers it as a favorite. For example, if the user checks the local keyword map and finds that multiple keywords of interest are displayed, the patent is registered as a favorite. Figure 11 shows an example of the screen display for registering a patent of interest as a favorite. When the user selects patent "JP51575" on the global patent map (step S404), the keywords contained in patent "JP51575" and patents located nearby are displayed on the local keyword map (step S405). The user checks the local keyword map, and if the user is interested in patent "JP51575," registers it as a favorite using the instruction unit 4 (Yes in S406). To register a patent as a favorite, the user selects patent "JP51575," for example, displays a button for registering as a favorite, and selects that button. When the user registers a patent as a favorite, patent "JP51575" and patents located nearby are registered (step S407).

[0046] The user can select the extent to which patents located near the selected patent are to be registered. Various methods are available, including registering patents that contain a preset percentage of the same keywords as patent "JP51575" or registering patents whose similarity between patents exceeds a preset value. The range of patents to be included in the favorites can be the same as the range of patents containing the keywords displayed in the local keyword map, or it can be a narrower or wider range, and can be set by the user. The registered patents are loaded into the integrated map generator 20 along with the registered keywords, and an integrated map covering the registered data is generated. The integrated map controller 25 controls the display unit 5 to display the generated integrated map, a patent my map viewed from the patent's perspective, and a keyword my map viewed from the keyword's perspective (step S408).

[0047] A user selects an arbitrary keyword from the global keyword map, checks patents related to the selected keyword on the local patent map, and if the user is interested in the selected keyword, registers it as a favorite. For example, when checking the local patent map, if multiple patents that interested the user in a previous operation are displayed, the keyword can be registered as a favorite. FIG. 12 shows an example of a screen display for registering a keyword of interest as a favorite. When the user selects the keyword "driving assistance" on the global keyword map (step S411), patents containing the keyword "driving assistance" and keywords located nearby are displayed on the local patent map (step S412). The user checks the local patent map, and if the user is interested in the keyword "driving assistance," registers it as a favorite using the instruction unit 4 (Yes in S413). To register as a favorite, the user selects the keyword "driving assistance," for example, displays a button for registering as a favorite, and selects that button. When the user registers as a favorite, the keyword "driving assistance" and keywords located nearby are registered (step S414).

[0048] The extent to which keywords located near the selected keyword are to be registered can be selected by the user as appropriate, similar to the selection of favorite patents. The registered keywords, along with the registered patents, are read into the integrated map generation unit 20, and an integrated map covering the registered data is generated. The integrated map control unit 25 controls the display unit 5 to display the generated integrated map as a patent my map and a keyword my map (step S415).

[0049] As shown in Fig. 13, the Patent My Map and Keyword My Map are displayed on two new screens in addition to the four screens displaying the global patent map, global keyword map, local keyword map, and local patent map, as a result of user operation via the instruction unit 4. Note that the Patent My Map and Keyword My Map may also be displayed on six screens at the same time, in addition to the four screens displaying the global patent map, global keyword map, local keyword map, and local patent map.

[0050] Unless an end instruction is given by the instruction unit 4 (No in step S418), the user repeats various operations on the global patent map, global keyword map, local keyword map, and local patent map. When the user gives an end instruction by the instruction unit 4 (Yes in step S418), the information retrieval device 1 ends.

[0051] In this way, users can register patents and keywords that interest them as favorites, and the patents and keywords of interest are added to their Patent My Map and Keyword My Map, updating each My Map. Users can understand their own areas of interest by checking their Patent My Map and Keyword My Map. If a user has filed a patent and registers that patent as a favorite on the Global Patent Map, they can check their Patent My Map to find patents related to their application, and check their Keyword My Map to find related keywords. Users can also check their My Map to understand the relationship between patents and keywords, understand the worldview, trends, and power relationships in their field, and understand the position of their patents in the overall picture. Based on this information, users can develop strategies for determining what patents they should file in the future.

[0052] Furthermore, once a Keyword My Map is created, it can be used to analyze other patent data by applying the created Keyword My Map to other patent data consisting of different populations.

[0053] In this embodiment, patents and keywords are registered in My Maps by the user performing a favorite registration operation, but the number of times the user selects the same patent or the same keyword on the global map can be stored as a history, and if the same patent or the same keyword is selected more than a preset number of times, the patent, nearby patents, their keywords, and their nearby keywords can be automatically registered in My Maps.

[0054] Another method for selecting an element on the global map is described below. Figure 14 shows the changes in the local keyword map screen display as elements are successively selected on the global patent map. As shown in Figure 14(a), the user uses the indicator 4 to select patent "JP51575" on the global patent map displayed on the display 5, and then gradually moves to the right while repeatedly selecting elements. Figures 14(b)-(d) show the changes in the local keyword map screen display at this time. When the user selects patent "JP51575" on the global patent map, the keywords contained in "JP51575" and patents located nearby are displayed on the local keyword map, as shown in Figure 14(b). As the user repeatedly selects elements while gradually moving to the right on the global patent map, the arrangement of the displayed keywords continuously changes, as shown in Figures 14(b)-(d). Keywords that are no longer included gradually become lighter and disappear, while newly included keywords gradually become darker and appear. Specifically, with successive operations on the global patent map, the position of the keyword "eyes off" continuously moves to the right, the display of the keyword "other vehicle detection" gradually fades and disappears, and the display of the keyword "collision avoidance" gradually fades from a non-displayed state to a displayed state.

[0055] The changes in the local keyword map that accompany successive operations on the global patent map have been explained above, but the same can be said for changes in the local patent map that accompany successive operations on the global keyword map. As the user performs successive operations on the global keyword map, the display on the local patent map also changes continuously.

[0056] By continuously selecting elements on the global map in this way, the screen display of the local map changes continuously in response to the operation, allowing you to continuously grasp the changes in related keywords as you operate on the patent you select, and the changes in related patents as you operate on the keyword you select, allowing you to accurately select patents and keywords that interest you.

[0057] [3. Effects, etc.] The information search device 1 in this embodiment is an information search device 1 that visualizes patent data consisting of patents and patent-related keywords on a map and searches for patents, and is equipped with an integrated map generation unit 20 that generates an integrated map by mapping a model estimated based on multiple different multidimensional perspectives of the patent data to a lower dimension, and an integrated map control unit 25 that controls the integrated map to display a global map for each perspective when viewed from the perspective of the patent and keyword, and when an arbitrary element on the global map viewed from the perspective of the displayed patent is selected, controls to display a local map viewed from the perspective of keywords related to the selected element and elements located in its vicinity, and when an arbitrary element on the global map viewed from the perspective of the displayed keyword is selected, controls to display a local map viewed from the perspective of the patent related to the selected element and elements located in its vicinity.

[0058] Users can find patents and keywords of interest by repeatedly operating and checking the global patent map, local keyword map, global keyword map, and local patent map in turn.

[0059] When any element on the global map is selected, the integrated map control unit 25 may allow the user to register that element and elements located in its vicinity at their instruction, and may control the display of a My Map for each perspective when viewed from each of the patents and keywords consisting of the registered elements.

[0060] Users can understand their areas of interest by checking their Patent My Map and Keyword My Map. Furthermore, by registering and checking the patents they have filed on My Map, they can find patents and keywords related to their own patents. By grasping the relationship between patents and keywords and understanding the worldview, trends, and power relationships in their field, they can grasp the overall position of their patents and develop strategies for future applications.

[0061] When any element on the global map is successively selected, the integrated map control unit 25 may control the display so that the local map is continuously changed when viewed from the perspective of search target information or key information related to the selected element and elements located in its vicinity.

[0062] By continuously operating on the global map, users can continuously grasp the changes in related keywords as they operate on the patents they select, and the changes in related patents as they operate on the keywords they select, allowing them to accurately select patents and keywords that interest them.

[0063] The integrated map control unit 25 may control the display of an element corresponding to the selected element on the global map from the patent's perspective when any element on the local map from the patent's perspective is selected, and may control the display of an element corresponding to the selected element on the global map from the keyword's perspective when any element on the local map from the keyword's perspective is selected.

[0064] This allows users to check the local map and easily select patents or keywords that interest them on the global map.

[0065] (Embodiment 2) As the second embodiment, a screen display method different from that of the first embodiment will be described. The configuration of the information retrieval device 1 is the same as that of the first embodiment, and therefore the description will be omitted. The search target information used in the second embodiment is the patent data shown in Fig. 2, the same as that of the first embodiment.

[0066] 15 is a flowchart showing the operation of the information retrieval device 1 in embodiment 2. The patent data shown in FIG. 2 is stored in the aggregated data storage unit 3. The integrated map generation unit 20 reads all of the patent data stored in the aggregated data storage unit 3 (step S1501) and generates an integrated map for all of the patent data (step S1502). The operation of generating the integrated map is the same as the operation described in embodiment 1.

[0067] In response to an instruction from the user via the instruction unit 4 to display the integrated map, the integrated map control unit 25 controls the display unit 5 to display the generated integrated map covering all patent data (step S1503). The display screen of the display unit 5 is shown in Figure 16. The display screen is composed of four screens: a global patent map, a global keyword map, a local patent map, and a local keyword map. The global map displays the integrated map covering all patent data. The global patent map is the integrated map covering all patent data viewed from the perspective of patents, and the global keyword map is the map viewed from the perspective of keywords. The local patent map is created by selecting any range on the global patent map and displaying the patents within the selected range. The local keyword map is the integrated map covering the patents displayed on the local patent map viewed from the perspective of keywords.

[0068] The user selects a range on the global map displayed on the display unit 5 (step S1504). FIG. 17 shows an example of the screen display when the user selects a range on the global patent map. The integrated map control unit 25 acquires a list of patents selected by the user on the global patent map on the display unit 5, and extracts patent data corresponding to the acquired list of patents from the aggregated data storage unit 3 (step S1505). The integrated map control unit 25 outputs the extracted patent data to the integrated map generation unit 20.

[0069] The integrated map generation unit 20 generates an integrated map for the extracted patent data (step S1506). The operation of generating an integrated map is the same as the operation of generating an integrated map for all patent data. The integrated map control unit 25 controls the display unit 5 to display the integrated map for the extracted patent data as a local patent map and a local keyword map (step S1507). Since the local patent map and the local keyword map are integrated maps modeled taking into account their respective viewpoints, a single-viewpoint map for one viewpoint is a map that takes into account other viewpoints. Furthermore, in each single-viewpoint map, elements with high similarity to each other are placed close to each other.

[0070] FIG. 18 shows an example of a screen display when a user selects a range on the global keyword map displayed on the display unit 5 (step S1506). The integrated map control unit 25 acquires a list of keywords selected by the user on the global keyword map on the display unit 5, and extracts patent data of patents containing the acquired keywords from the aggregated data storage unit 3 (step S1505). The integrated map control unit 25 outputs the extracted patent data to the integrated map generation unit 20. The integrated map generation unit 20 generates an integrated map for the extracted patent data (step S1506). The integrated map control unit 25 controls the integrated map for the extracted patent data to be displayed on the display unit 5 as a local patent map and a local keyword map (step S1507).

[0071] In this way, the local patent map and local keyword map select patents and keywords within a certain range and generate and display an integrated map, making it possible to grasp patents, keywords, and their relationships that could not be grasped with a global map that covers all patents.

[0072] In the second embodiment, a form in which a global map and a local map are simultaneously displayed on the display unit 5 has been described, but a form in which the global map and the local map are sequentially switched and displayed may also be used.

[0073] Next, the user selects any element on the local map (step S1508). Figure 19 shows the screen display when the user selects any patent on the local patent map. In Figure 19, when the user selects a patent on the local patent map, the patent on the global patent map that is the same as the selected patent is explicitly displayed (step S1509). To explicitly display a patent on the global patent map, the integrated map control unit 25 changes the background color of the relevant patent. The explicit display method is not limited to this method, and may also be a method of flashing the relevant patent, changing the font color of the patent number, or the like.

[0074] In this embodiment, the user selects one patent when selecting an arbitrary patent, but multiple patents may be selected on the local patent map. In this case, the multiple corresponding patents are explicitly displayed on the global patent map.

[0075] FIG. 20 shows the screen display when an arbitrary keyword is selected on the local keyword map. In FIG. 20, when the user selects a keyword on the local keyword map, the same keyword on the global keyword map is explicitly displayed (step S1509). To explicitly display a keyword on the global keyword map, the integrated map control unit 25 changes the background color of the corresponding keyword. The explicit display method is not limited to this method, and may also be a method of blinking the corresponding keyword or a method of changing the character color of the keyword. Multiple keywords may also be selected on the local keyword map. In this case, the multiple corresponding keywords are explicitly displayed on the global keyword map.

[0076] Unless the user issues an instruction to end the process using the instruction unit 4 (No in step S1510), the process of selecting an arbitrary range on the global map (step S1504) and explicitly displaying the elements selected on the local map on the global map (step S1510) is repeated.

[0077] In this way, the information retrieval device 1 provides a global map displaying all patents and keywords targeted by the search, enabling visual search and exploration. Furthermore, the information retrieval device 1 generates and visualizes a local map specialized for a selected range, enabling detailed search and exploration. When a target is identified on the local map, the information retrieval device 1 displays the corresponding point on the global map. By sequentially referring to the global map and local map displays while identifying the target, users can broadly search for patents in the area of ​​interest, narrowing their search through operations to find the patent of interest. Furthermore, even if users do not know the keywords required for the search, they can understand the relationships between patents, keywords, and patents and keywords through operations, thereby finding a group of patents of interest. Furthermore, not only can users find a group of patents of interest, but they can also visualize and understand their interrelationships, thereby gaining knowledge in the field.

[0078] (Other embodiments) As described above, the embodiment has been described as an example of the present invention. However, the present invention is not limited to this, and can be applied to embodiments in which modifications, substitutions, additions, omissions, etc. are made. Therefore, other embodiments will be exemplified below.

[0079] In the first and second embodiments, patents are used as search target information. Another embodiment will be described, in which papers are used as search target information. When researchers venture into a new research field, they often do not know the research trends and keywords in that field. Using the information retrieval device 1 described in the first embodiment, the researcher creates an integrated map of paper data consisting of papers and keywords. The researcher selects a paper from the global paper map, checks keywords related to the selected paper from the local keyword map, selects a keyword from the global keyword map, and checks papers related to the selected keyword from the local paper map. The researcher repeats this process, registering papers and keywords of interest as favorites. By checking the My Paper Map, which consists of favorited papers, and the My Keyword Map, which consists of favorited keywords, the researcher can identify the papers and keywords of interest to them and understand the worldview and trends in that field. Based on this information, the researcher can develop strategies for future research.

[0080] Another embodiment will now be described, which involves collecting information about travel destinations. For example, when traveling to Kyushu for the first time, you may not know where things are or what keywords to use when searching. An integrated map is created using the information search device 1 described in the first embodiment, targeting travel data consisting of keywords for sightseeing spots and delicious restaurants for each location in Kyushu. Travelers repeatedly operate the global location map, global keyword map, local location map, and local keyword map to register places and keywords of interest to their favorites. By referring to the registered My Maps, travelers can identify places and keywords that interest them and are likely to be potential travel destinations, which can be used in planning their trips. Once a Keyword My Map of interest to a travel destination is created, the Keyword My Map can be applied to other travel destinations (for example, travel data consisting of keywords for sightseeing spots and delicious restaurants for each location in Hokkaido) and used in planning other travel destinations.

[0081] Furthermore, as another embodiment, a form of searching for information on a social networking service (SNS) will be described. Posts on the SNS are tagged with classification tags called hashtags. A user creates an integrated map using the information retrieval device 1 described in the first embodiment, targeting posts as search target information and hashtags as key information, and repeatedly operates the global post map, global hashtag map, local post map, and local hashtag map to register posts and hashtags of interest to their favorites. By referring to their favorite My Maps, the user can identify posts and hashtags of interest to them. [Explanation of symbols]

[0082] 1 Information retrieval device 2 Integrated Map Section 3. Aggregated data storage unit 4 Instruction section 5 Display section 20 Integrated map generation unit 21 Initialization processing section 22 Individual map information generation unit 23 Placement position calculation section 24 Integrated map information generation unit 25 Integrated map control unit

Claims

1. An information search device that visualizes search data consisting of search target information and key information related to the search target information on a map and searches for the search target information, an integrated map generation unit that generates an integrated map by mapping a model estimated based on a plurality of different multidimensional viewpoints of the search data onto a lower dimension; an integrated map control unit that controls the integrated map to display a global map for each viewpoint when viewed from each viewpoint of the search target information and the key information, and when an arbitrary element on the displayed global map is selected, controls to display a local map when viewed from the viewpoint of the search target information or the key information regarding the selected element and elements located nearby; An information search device comprising:

2. When an arbitrary element on the global map is selected, the integrated map control unit allows the user to register the element and elements located nearby, and controls the display of the search target information consisting of the registered elements and My Maps for each viewpoint when viewed from each viewpoint of the key information.

2. The information retrieval device according to claim 1.

3. the integrated map control unit controls, when any element on the global map is successively selected, to continuously change and display a local map viewed from the viewpoint of the search target information or the key information related to the selected element and elements located nearby; 2. The information retrieval device according to claim 1.

4. the integrated map control unit, when an arbitrary element on the local map seen from the viewpoint of the search target information is selected, controls to explicitly display an element on the global map seen from the viewpoint of the search target information corresponding to the selected element, and when an arbitrary element on the local map seen from the viewpoint of the key information is selected, controls to explicitly display an element on the global map seen from the viewpoint of the key information corresponding to the selected element.

3. The information retrieval device according to claim 1.

5. the integrated map control unit, when an arbitrary element on the global map seen from the viewpoint of the search target information is selected, controls to display a local map seen from the viewpoint of the key information relating to the selected element and elements located nearby, and when an arbitrary element on the global map seen from the viewpoint of the key information is selected, controls to display a local map seen from the viewpoint of the search target information relating to the selected element and elements located nearby.

2. The information retrieval device according to claim 1.

6. An information retrieval method for visualizing search data consisting of search target information and key information related to the search target information on a map, and retrieving the search target information, comprising: an integrated map generation step of generating an integrated map by mapping a model estimated based on a plurality of different multidimensional viewpoints of the search data onto a lower dimension; an integrated map control step of controlling the integrated map to display a global map for each viewpoint when viewed from each viewpoint of the search target information and the key information, and when an arbitrary element on the displayed global map is selected, controlling to display a local map when viewed from the viewpoint of the search target information or the key information regarding the selected element and elements located nearby; An information retrieval method comprising:

7. An information retrieval program that causes a computer to execute an information retrieval method for visualizing search data, which is composed of search target information and key information related to the search target information, on a map, and retrieving the search target information, an integrated map generation step of generating an integrated map by mapping a model estimated based on a plurality of different multidimensional viewpoints of the search data onto a lower dimension; an integrated map control step of controlling the integrated map to display a global map for each viewpoint when viewed from each viewpoint of the search target information and the key information, and when an arbitrary element on the displayed global map is selected, controlling to display a local map when viewed from the viewpoint of the search target information or the key information regarding the selected element and elements located nearby; An information retrieval program that executes the above.

Images