JP6332955B2 - Map display method and map display device - Google Patents

Description

  The present invention relates to a map display method and a map display apparatus for displaying the closeness of contents between documents.

  A service for searching for documents with similar contents from a large number of documents is provided (for example, see Patent Document 1). Patent Document 1 plots the points of each document so that similar documents are arranged nearby. As a result, a map representing the similarity between the documents can be created.

WO2008 / 143116

  There are many cases in which it is necessary to compare the flow of information such as technological trends, fashion trends, and sensor measurements. In such a case, since the map of Patent Document 1 needs to be created for the extracted set after being extracted at the time, the same operation has to be repeated for the number of times to be compared. Moreover, since the user needs to set the extraction time for the map of Patent Document 1, the user is required to know the subject to be compared. For this reason, it is not easy to compare and examine the fluidity of temporal information using the map of Patent Document 1.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a map display method and a map display system that make it easy to make a comparative examination of temporal information fluidity.

  The present invention highlights points that coincide with the time designated by the user while the map is displayed.

The map display method of the present invention is
A data group including a plurality of unit data including vector information that can be vectorized different from the time information and including the time information in the meta information is vectorized for each unit data based on the vector information, and the point represented by the vector is A mapping procedure to place on the map that represents the proximity of the vector;
An extraction procedure for extracting unit data corresponding to the set time information from the data group;
An emphasis procedure for emphasizing the points of the extracted unit data among the points of each unit data arranged on the map;
Have

In the map display method of the present invention,
A time scale creation procedure for assigning time information included in the data group to time scale graduations in time series order is further included between the mapping procedure and the extraction procedure,
In the extraction procedure, unit data corresponding to time information set on the scale of the time scale may be extracted from the data group.

In the map display method of the present invention,
In the extraction procedure, the number of extracted unit data is calculated,
In the enhancement procedure, the calculated number may be arranged in the vicinity of the scale.

In the map display method of the present invention,
The set time information in the extraction procedure is sequentially set in chronological order,
In the enhancement procedure, the extracted unit data may be sequentially enhanced in time series order.

In the map display method of the present invention,
The data group has one or more meta information common to each unit data,
In the extraction procedure, at least one value of the meta information is further set, and unit data corresponding to the set time information and corresponding to the set meta information value is extracted from the data group May be.

In the map display method of the present invention,
The data group has one or more meta information common to each unit data,
In the extraction procedure, for the set meta information of the meta information, calculate the appearance frequency in the set formed by the extracted unit data,
In the emphasis procedure, the set meta information value of the specific appearance frequency may be arranged in the vicinity of the map.

The map display system of the present invention is
A group of data that includes multiple unit data that includes vector information that can be vectorized differently from the time information, including time information in the meta information, is vectorized for each unit data based on information different from the time information, and is represented by a vector. A mapping unit that arranges points on the map representing the proximity of vectors,
An extraction unit for extracting unit data corresponding to the set time information from the data group;
An emphasizing unit for emphasizing the points of the extracted unit data among the points of each unit data arranged on the map;
Is provided.

In the map display system of the present invention,
A time scale creation unit that assigns the time information included in the data group to the scale of the time scale in chronological order;
The extraction unit may extract unit data corresponding to time information set on the scale of the time scale from the data group.

In the map display system of the present invention,
The extraction unit calculates the number of extracted unit data,
The enhancement unit may arrange the calculated number in the vicinity of the scale.

In the map display system of the present invention,
The set time information in the extraction unit is sequentially set in chronological order,
The enhancement unit may sequentially enhance the extracted unit data in chronological order.

In the map display system of the present invention,
The data group has one or more meta information common to each unit data,
The extraction unit further sets at least one value of the meta information, and extracts unit data corresponding to the set time information and corresponding to the set meta information value from the data group May be.

In the map display system of the present invention,
The data group has one or more meta information common to each unit data,
The extraction unit calculates the appearance frequency in the set formed by the extracted unit data for the set meta information of the meta information,
The emphasizing unit may arrange meta information values of a set specific appearance frequency in the vicinity of the map.

  The above inventions can be combined as much as possible.

  ADVANTAGE OF THE INVENTION According to this invention, the map display method and map display system which make easy the comparative examination of the fluidity | liquidity of time information can be provided.

1 shows an example of a configuration of a map display system according to an embodiment of the present invention. The sequence of the map display system which concerns on embodiment of this invention is shown. An example of a data group when unit data is a public patent gazette is shown. An example of mapping is shown. The 1st example of a display on a display part is shown. The 2nd example of a display on a display part is shown.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, this invention is not limited to embodiment shown below. These embodiments are merely examples, and the present invention can be implemented in various modifications and improvements based on the knowledge of those skilled in the art. In the present specification and drawings, the same reference numerals denote the same components.

(Embodiment 1)
FIG. 1 shows a configuration example of a map display system according to the present embodiment. The map display system according to the present embodiment includes a server 10, a storage 20, and a user terminal 30. The storage 20 is an arbitrary storage medium accessible from the server 10. Any number of servers 10, storages 20, and user terminals 30 may be employed. In the present embodiment, a case where there is one server 10, two storages 20, and one user terminal 30 will be described.

  FIG. 2 shows a sequence of the map display system according to the present embodiment. The storage 20 holds a plurality of unit data. The unit data includes time information and vector information that can be vectorized. The user terminal 30 transmits a mapping request via the communication network (S101). Upon receiving a mapping request from the user terminal 30, the server 10 acquires a plurality of unit data from the storage 20 (S102), creates a requested map using the acquired unit data (S103), and creates a map. It transmits to the user terminal 30 (S104). The user terminal 30 displays the map received from the server 10 (S105).

  When the time is set from the user terminal 30 (S106), the server 10 extracts unit data corresponding to the set time information in the map (S107) and emphasizes the extracted unit data (S108). The new map is transmitted to the user terminal 30 (S109). The user terminal 30 displays the map received from the server 10 (S110).

  Any information that can be vectorized can be mapped. The present invention can be applied to arbitrary information having time information and vector information that can be vectorized. In the present embodiment, as an example, a case where the unit data is an open patent publication will be described. The published patent gazette includes vector information such as claims, detailed explanation, and summary, as well as time information such as filing date, publication date, and priority date. Since the claims, the detailed description, and the summary include words, a document vector having the words as elements can be generated for each published patent publication.

FIG. 3 shows an example of a published patent publication according to this embodiment. In October 2010, Company A filed application has been published in the publication number PN _A, application B, Inc., filed in May 2004 has been published in publication number PN _B, C Company was filed on 10 May 2011 application has been published in the publication number PN _C, D, Inc. is the application filed has been published in the publication number PN _D in October 2003.

  The server 10 includes a communication function unit (not shown) that transmits / receives data to / from the user terminal 30 and the storage 20 via a communication network, and a configuration for creating a map. The configuration for creating a map includes, for example, a mapping unit 11, a time scale creation unit 17, an extraction unit 15, and an enhancement unit 16.

  When creating the map, the server 10 executes the map display method according to the present embodiment. The map display method according to the present embodiment includes a mapping procedure, a time scale creation procedure, an extraction procedure, and an enhancement procedure in this order. The mapping procedure and time scale creation procedure are performed in step S103 shown in FIG. 2, the extraction procedure is performed in step S107 shown in FIG. 2, and the enhancement procedure is performed in step S108 shown in FIG.

  In the mapping procedure, the mapping unit 11 vectorizes a data group having a plurality of unit data for each unit data based on the vector information, and places the points represented by the vector on a map representing the proximity of the vector To do. The vector information is, for example, a character string described in the detailed description. Since the character string described in the detailed description includes a word, vectorization using the word as an element is possible.

FIG. 4 shows an arrangement example on the map. Each point on the map represents each published patent publication. The closer the published patent gazette of the invention content is, the closer it is arranged. By vectorizing each patent publication based on the character string described in the detailed description, on the map that represents the closeness of the invention described in the detailed description, placing the points of PN _A to PN _D can do. The mapping method is arbitrary, but for example, a vector space model in a concept search algorithm can be used.

  In the time scale creation procedure, the time scale creation unit 17 assigns the time information included in the data group to be mapped to the scales of the time scale in chronological order. Although the shape of the time scale is arbitrary, for example, the application dates are assigned in chronological order on one axis as shown in FIG. At this time, one scale may indicate only one day of the filing date, or may indicate a period such as January 1, 2003 to December 31, 2003.

  The server 10 transmits a screen display in which a time scale is arranged in the vicinity of the map shown in FIG. 4 to the user terminal 30 (S104). Then, the user terminal 30 displays a time scale together with the map shown in FIG. 4 (S105).

One of the time scales is set and set from the user terminal 30 (S106), and the memory indicates the filing date from July 10, 2003 to December 5, 2004. Then, the server 10 executes an extraction procedure (S107). In the extraction procedure, the extraction unit 15 extracts unit data corresponding to the set time information from the data group. Extractor 15, filing date extracts published patent publications Publication No. PN _B and PN _D contained 10 July - December 5, 2004, 2003. Then, in the emphasis steps (S108), emphasizing section 16 emphasizes the _{P B} point indicating the patent publication of a public number PN _B, and _{P D} point indicating the patent publication of a public number PN _D, a. This makes it possible to create a new map highlighting points corresponding to the set time information in the map. For example, as shown in FIG. 5, the periphery of the point is surrounded by a color different from the point. Then, steps S109 and S110 are performed. Thereby, the map shown in FIG. 5 can be displayed on the user terminal 30.

Further, it is assumed that one of the memories in the time scale is set and set from the user terminal 30 (S106), and that memory indicates the application date from July 28, 2010 to December 24, 2011. Then, the server 10 executes an extraction procedure (S107). The extraction procedure, extraction section 15, filing date extracts published patent publications Publication No. PN _A and PN _C included in the July 28, 2011 to December 24, 2010. Then, in the emphasis steps (S108), enhancement unit 16, as shown in FIG. 6, a point _{P A} indicating the patent publication of a public number PN _A, and _{P C} point indicating the patent publication of a public number PN _C To emphasize. Then, steps S109 and S110 are performed. Thereby, the map shown in FIG. 6 can be displayed on the user terminal 30.

  As described above, the server 10 transmits the map shown in FIGS. 5 and 6 to the user terminal 30 (S109), and the user terminal 30 displays the map shown in FIGS. 5 and 6 (S110). The maps shown in FIG. 5 and FIG. 6 are emphasized only for the points corresponding to the set time, while the maps representing the similarity of the invention are not changed. For this reason, the map display system according to the present embodiment can easily perform a comparative study of temporal information fluidity.

  Note that the time information set in the extraction procedure may be automatically and sequentially set in chronological order. For example, an automatic playback button is arranged in the vicinity of the map transmitted to the user terminal 30 in step S104. When the automatic reproduction button is pressed on the user terminal 30, the user terminal 30 transmits an automatic reproduction instruction to the server 10. When the extraction unit 15 acquires the automatic reproduction instruction, the extraction unit 15 automatically sets the time information in chronological order. The extraction unit 15 sequentially extracts the published patent gazettes corresponding to the set time information from the data group while sequentially changing the set time information in chronological order. And the emphasis part 15 emphasizes sequentially the published patent gazette group extracted for every time information, and animates it in time series. Thereby, the transition of the distribution according to the time transition of the published patent publication can be displayed.

  In the extraction procedure, the extraction unit 15 preferably calculates the number of extracted unit data. For example, it is assumed that the number of published patent gazettes included in the application date from July 10, 2003 to December 5, 2004 is M. In this case, in the enhancement procedure, the enhancement unit 16 arranges the calculated number “M” in the vicinity of the scale. Thereby, as shown in FIG. 5, the number of published patent gazettes corresponding to the set time can be displayed as “M” in the vicinity of the scale.

  The data group may have one or more pieces of meta information common to each unit data. For example, each published patent publication has common meta information called an applicant. In step S106 shown in FIG. 2, when there is a setting for calculating the appearance frequency of the applicant from the user terminal 30, in the extraction procedure (step S107), the extraction unit 15 is formed by the extracted published patent publication. The frequency of occurrence of the applicant's value within the set is calculated. At this time, for example, when the appearance frequency of company B is the highest and the appearance frequency of company D is the next highest, the company names of company B and company D are arranged in the vicinity of the map as shown in FIG.

In the present embodiment, the data group may have one or more pieces of meta information common to each unit data. For example, each published patent publication has common meta information called an applicant. In step S106 shown in FIG. 2, it is assumed that an application date from July 10, 2003 to December 5, 2004 is set from the user terminal 30 and the applicant is set to B. In this case, the extraction procedure (step S107), extraction unit 15, filing date of the publication number PN _B and PN _D contained 10 July - December 5, 2004, 2003, further applicant extract the P _B point indicating the patent publication of publication number PN _B is B company, P _D point indicating the patent publication of publication number PN _D is not extracted. Then, the enhancement procedure, emphasis section 16 emphasizes the P _B point indicating the patent publication of a public number PN _B, P _D point indicating the patent publication of publication number PN _D is not stressed.

  In the present embodiment, the case where the unit data is an open patent publication has been described as an example, but the present invention is not limited to this. For example, it can be applied to arbitrary information including time information and vector information instead of the published patent publication.

(Embodiment 2)
In the present embodiment, details of a method for arranging vector information on a map using a vector space model in a concept search algorithm in a mapping procedure will be described. Also in this embodiment, the case where vector information is a character string having a word as an element will be described as an example.

When the unit data d _i can be expressed in matrix with respect to the element t _j , the unit data d _i can be described by the vector space model d _i = (t ₁ , t ₂ , t ₃ ,...). Unit data _{d i} is the document _{d i,} is an element _{t i} is a word _{t j,} when the word frequency _{t j} appearing in the document _{d i} and _{n ij,} document _{d i} is the document vector _d i = ( n _i1 , n _i2 , n _i3 ,...).

Word frequencies _{t 1,} t 2, _{t 3} in the document _{d 1} is 0,1,0, respectively, word frequency _{t 1,} t 2, _{t 3} in the document _{d 2} is respectively 2,1,0 Yes, when the appearance frequencies of the words t ₁ , t ₂ , and t ₃ in the document d ₃ are ₁ , ₂ , and ₃ , respectively, the word document matrix M is expressed as follows.

In the document set d = Σd _i , the proximity between the document vectors of each document d _i is calculated. When the contents of the document are close, the types of words used are similar, and the points indicated by the document vector are arranged close to each other. By calculating the closeness between the document vectors, the closeness of the contents of the document can be obtained. The calculation may be a distance between document vectors, or an arbitrary calculation such as inner product or outer product.

  Here, words commonly used in any document do not affect the proximity of the content of the document. Therefore, in the calculation of the document vector, it is preferable to provide a difference in the contribution of the word characteristic to each document and the other words to the document vector. For example, weighting is performed using a tfidf (Term Frequency Inverse Document Frequency) method. Thereby, the precision of the closeness of the content of a document can be improved.

  Based on the obtained proximity, mapping is performed by a mapping algorithm. For example, the point indicated by the document vector is arranged on the surface. Thereby, document vectors of all documents can be plotted in a multidimensional vector space. At this time, it is preferable to arrange the documents whose semantic distances are separated with reduced accuracy of the distance between the planes of the documents. Moreover, although it may arrange | position on a plane as shown in FIG. 2, you may arrange | position on a spherical surface.

  In the present embodiment, the case where the vector information is a character string including a word has been described, but the present invention is not limited to a document.

  For example, the vector information may be an arbitrary document. A document is an arbitrary medium that can identify words, and includes character data, image data such as moving images and still images, and audio data. Note that character data includes numbers and symbols, and words also include numbers and symbols.

For example, the user access log information in the online service can be exemplified as the vector information. The unit data d _i is the time, the element t _j is the user, and the number of accesses of the user t _j between the times d _{i to} d _i + T (time interval T) is n _ij . In this case, the time d _i can be expressed as a vector d _i = (n _i1 , n _i2 , n _i3 ,...). Thereby, mapping can be performed by the mapping algorithm.

For example, sensor information can be exemplified as vector information. The unit data d _i is the time, the element t _j is the sensor, and the output numerical value of the sensor t _j between the times d _{i to} d _i + T (time interval T) is n _ij . In this case, the time d _i can be expressed as a vector d _i = (n _i1 , n _i2 , n _i3 ,...). Thereby, mapping can be performed by the mapping algorithm.

For example, image information can be exemplified as vector information. The unit data d _i is an image, the element t _j is a frequency component, and the numerical value of the frequency component t _j is n _ij . Components here frequency, the image d _i and the frequency conversion is a component of each frequency after conversion. In this case, the image d _i can be expressed as a vector d _i = (n _i1 , n _i2 , n _i3 ,...). Thereby, mapping can be performed by the mapping algorithm.

  As described above, by using the vector space model in the concept search algorithm, unit data having vector information can be arranged on the map.

  The present invention can be applied to the information communication industry.

10: Server 11: Mapping unit 15: Extraction unit 16: Emphasis unit 17: Time scale creation unit 20: Storage 30: User terminal

Claims (12)

A data group including a plurality of unit data including vector information that can be vectorized different from the time information and including the time information in the meta information is vectorized for each unit data based on the vector information, and the point represented by the vector is Mapping procedure to place on the map according to the closeness of the contents of the vector information,
An extraction procedure for extracting unit data corresponding to the set time information from the data group;
An emphasis procedure for emphasizing the points of the extracted unit data among the points of each unit data arranged on the map;
A map display method. A time scale creation procedure for assigning time information included in the data group to time scale graduations in time series order is further included between the mapping procedure and the extraction procedure,
The map display method according to claim 1, wherein in the extraction procedure, unit data corresponding to time information set on the scale of the time scale is extracted from the data group. In the extraction procedure, the number of extracted unit data is calculated,
The map display method according to claim 2, wherein the calculated number is arranged in the vicinity of the scale in the emphasis procedure. The set time information in the extraction procedure is sequentially set in chronological order,
The map display method according to any one of claims 1 to 3, wherein in the emphasis procedure, the extracted unit data is sequentially emphasized in time series. The data group has one or more meta information common to each unit data,
In the extraction procedure, at least one value of the meta information is further set, and unit data corresponding to the set time information and corresponding to the set meta information value is extracted from the data group The map display method according to any one of claims 1 to 4, wherein: The data group has one or more meta information common to each unit data,
In the extraction procedure, for the set meta information of the meta information, calculate the appearance frequency in the set formed by the extracted unit data,
The map display method according to any one of claims 1 to 5, wherein, in the emphasis procedure, a value of meta information having a set specific appearance frequency is arranged in the vicinity of the map. A group of data that includes multiple unit data that includes vector information that can be vectorized differently from the time information, including time information in the meta information, is vectorized for each unit data based on information different from the time information, and is represented by a vector. A mapping unit that arranges points on the map representing the proximity of vectors,
An extraction unit for extracting unit data corresponding to the set time information from the data group;
An emphasizing unit for emphasizing the points of the extracted unit data among the points of each unit data arranged on the map;
A map display system comprising: A time scale creation unit that assigns the time information included in the data group to the scale of the time scale in chronological order;
The map display system according to claim 7, wherein the extraction unit extracts unit data corresponding to time information set on the scale of the time scale from the data group. The extraction unit calculates the number of extracted unit data,
The map display system according to claim 8, wherein the highlighting unit arranges the calculated number in the vicinity of the scale. The set time information in the extraction unit is sequentially set in chronological order,
The map display system according to any one of claims 7 to 9, wherein the emphasizing unit sequentially emphasizes the extracted unit data in time series order. The data group has one or more meta information common to each unit data,
The extraction unit further sets at least one value of the meta information, and extracts unit data corresponding to the set time information and corresponding to the set meta information value from the data group The map display system according to any one of claims 7 to 10, wherein: The data group has one or more meta information common to each unit data,
The extraction unit calculates the appearance frequency in the set formed by the extracted unit data for the set meta information of the meta information,
The map display system according to any one of claims 7 to 11, wherein the emphasizing unit arranges a value of meta information having a set specific appearance frequency in the vicinity of the map.

Images