Patents by Inventor Issei Yoshida
Issei Yoshida has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20130332432Abstract: Embodiments are directed to, for a database comprising a set of transactions, with each transaction including an itemset comprising one or more individual items, obtaining a set of all closed frequent itemsets by checking frequencies of transaction sets, wherein a given itemset, I, is frequent where at least some threshold number (?) of transactions contain I, and wherein the given itemset I is closed wherein it is impossible to add any item to I without changing the frequency thereof.Type: ApplicationFiled: July 18, 2012Publication date: December 12, 2013Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventor: Issei Yoshida
-
Publication number: 20130332431Abstract: Embodiments are directed to, for a database comprising a set of transactions, with each transaction including an itemset comprising one or more individual items, obtaining a set of all closed frequent itemsets by checking frequencies of transaction sets, wherein a given itemset, I, is frequent where at least some threshold number (?) of transactions contain I, and wherein the given itemset I is closed wherein it is impossible to add any item to I without changing the frequency thereof.Type: ApplicationFiled: June 12, 2012Publication date: December 12, 2013Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventor: Issei Yoshida
-
Patent number: 8190613Abstract: A computer implemented method for creating indices for a database having a plurality of documents each being associated with one or more keywords. The method includes the steps of: dividing the database into a plurality of subsets; separating the keywords into a plurality of keyword groups based upon modulo G of the hash value of the keyword for each subset; reading each document of each subset to create a first sub-index and writing same to a storage device of the computer; reading the first sub-indices to merge the first sub-indices into a second sub-index for each keyword group to write same to the storage device; and reading the second sub-indices from the storage device to merge the second sub-indices into an index for the database and write same on the storage device. A program and a system for creating indices are also provided.Type: GrantFiled: June 3, 2008Date of Patent: May 29, 2012Assignee: International Business Machines CorporationInventors: Daisuke Takuma, Issei Yoshida
-
Publication number: 20120096031Abstract: A system and method for enabling efficient extraction of only meaningful frequent itemsets. The system includes a decision unit that decides a new itemset that becomes an investigation target in the same sequence as that of searching an itemset tree in a depth-first manner and in descending order, a frequent occurrence determining unit that registers the frequency of occurrence of the new itemset in a table if the frequency of occurrence is equal to or more than a predetermined threshold, a correlation determining unit that determines whether there is a correlation between each item in the new itemset and a subset of remaining items that were removed from the new itemset, and a registration unit that registers the new itemset in a set of meaningful frequent itemsets if the determination is positive for all items of the new itemset.Type: ApplicationFiled: October 5, 2011Publication date: April 19, 2012Applicant: International Business Machines CorporationInventor: Issei Yoshida
-
Patent number: 8140546Abstract: A computer system, methods, and programs for creating an index for aggregating data in at least one tree structure including at least one node each including one label indicating node type and values. The system includes a node ID assignment processing unit for assigning IDs to the nodes in a post order; first, second, and third index creation processing units. The first unit creates a first index having one or more sets of data including the node ID and values included in the node; the second unit creates a second index having one or more sets of data including node ID and ID of a descendant node having the minimum ID; and the third unit creates a third index having one or more sets of data including IDs of one or more nodes having specific values.Type: GrantFiled: June 4, 2009Date of Patent: March 20, 2012Assignee: International Business Machines CorporationInventor: Issei Yoshida
-
Patent number: 8095652Abstract: An analysis system, information processing apparatus, activity analysis method, and program for analyzing activities of an information source on a network. The system and apparatus include an attribute extraction block for extracting, an information propagation graph acquisition block for searching action history data, and a characteristic user calculation block for calculating an amount characteristic. The method and program product include the steps of extracting an information characteristic value, searching action history data, registering the information, calculating an amount characteristic, and integrating the amount characteristic.Type: GrantFiled: February 27, 2009Date of Patent: January 10, 2012Assignee: International Business Machines CorporationInventors: Raymond Harry Putra Rudy, Akiko Suzuki, Hiroki Yanagisawa, Issei Yoshida
-
Patent number: 7945552Abstract: A system of the present invention stores: a first index which designates lists of keywords contained in texts from identifications of the respective texts; a second index which designates lists of texts containing keywords from identifications of the respective keywords; and the number of texts containing the respective keywords. Then, upon receiving an input of a text search condition, the system calculates an estimation of search time by the first index and an estimation of search time by the second index, and determines which one of the first and second indexes makes a search faster. Then, by using the index which has been determined to make the search faster, the system searches for keywords which appear in texts satisfying the text search condition with higher frequency.Type: GrantFiled: March 26, 2008Date of Patent: May 17, 2011Assignee: International Business Machines CorporationInventors: Daisuke Takuma, Issei Yoshida, Yuta Tsuboi
-
Patent number: 7792761Abstract: Embodiments herein properly determine relevance between persons. Embodiments herein provide a system for evaluating relevance between persons, including relevance detecting means for detecting directed relevance of a person as a target to another person based on a text in which evaluations or feelings of the other person by another person is expressed; storing means for storing the detected relevance; evaluating means for evaluating, with respect to a set of a first person and second person to be evaluated, relevance between the first person and the second person based on a plurality of relevances including the stored relevances and reaching the second person from the first person via at least one further person; and outputting means for outputting the relevance evaluated.Type: GrantFiled: October 24, 2006Date of Patent: September 7, 2010Assignee: International Business Machines CorporationInventors: Akiko Murakami, Yuta Tsuboi, Hideo Watanabe, Issei Yoshida
-
Publication number: 20100138442Abstract: To provide an information processing apparatus, a database system, an information processing method, and a program which ensure efficient database accesses by partitioning.Type: ApplicationFiled: November 6, 2009Publication date: June 3, 2010Applicant: International Business Machines CorporationInventors: Kaoru Shinkawa, Issei Yoshida
-
Publication number: 20090307214Abstract: A computer system, methods, and programs for creating an index for aggregating data in at least one tree structure including at least one node each including one label indicating node type and values. The system includes a node ID assignment processing unit for assigning IDs to the nodes in a post order; first, second, and third index creation processing units. The first unit creates a first index having one or more sets of data including the node ID and values included in the node; the second unit creates a second index having one or more sets of data including node ID and ID of a descendant node having the minimum ID; and the third unit creates a third index having one or more sets of data including IDs of one or more nodes having specific values.Type: ApplicationFiled: June 4, 2009Publication date: December 10, 2009Inventor: Issei Yoshida
-
Publication number: 20090222557Abstract: An analysis system, information processing apparatus, activity analysis method, and program for analyzing activities of an information source on a network. The system and apparatus include an attribute extraction block for extracting, an information propagation graph acquisition block for searching action history data, and a characteristic user calculation block for calculating an amount characteristic. The method and program product include the steps of extracting an information characteristic value, searching action history data, registering the information, calculating an amount characteristic, and integrating the amount characteristic.Type: ApplicationFiled: February 27, 2009Publication date: September 3, 2009Inventors: Raymond Harry Putra Rudy, Akiko Suzuki, Hiroki Yanagisawa, Issei Yoshida
-
Patent number: 7584187Abstract: Obtaining a closed pattern in a second itemspace which is made by adding a new item to a first itemspace based on a closed pattern obtained for the first itemspace. To provide a system for detecting a closed pattern, which is a maximum set of items, based on history information which records a set of items that is a target of transaction processing for each transaction. The system records at least a first closed pattern obtained for a first itemspace. The system generates an additional pattern by adding an additional item that is not included in the first itemspace and compares a set of transactions including the whole of the additional pattern with a set of transactions including any of the whole of the first recorded closed pattern. The system detects union of the first closed pattern and the additional pattern as a second closed pattern in the second itemspace, on condition that the compared sets are identical with each other.Type: GrantFiled: July 5, 2006Date of Patent: September 1, 2009Assignee: International Business Machines CorporationInventor: Issei Yoshida
-
Patent number: 7584184Abstract: A system of the present invention stores: a first index which designates lists of keywords contained in texts from identifications of the respective texts; a second index which designates lists of texts containing keywords from identifications of the respective keywords; and the number of texts containing the respective keywords. Then, upon receiving an input of a text search condition, the system calculates an estimation of search time by the first index and an estimation of search time by the second index, and determines which one of the first and second indexes makes a search faster. Then, by using the index which has been determined to make the search faster, the system searches for keywords which appear in texts satisfying the text search condition with higher frequency.Type: GrantFiled: November 2, 2006Date of Patent: September 1, 2009Assignee: International Business Machines CorporationInventors: Daisuke Takuma, Issei Yoshida, Yuta Tsuboi
-
Publication number: 20090157659Abstract: Systems, methods, and computer products for information sharing using personalized index caching. Exemplary embodiments include a method including receiving a search query history from a node X in a node A, extracting characteristics of an index of the node A, searching the extracted characteristics, which include a file ID that is included in the index of the node A, adding metadata information to the index of the node A, in response to a determination that the node A includes at least one additional local metaindex, searching the at least one additional metaindex with the search query history from the node X in the node A, and merging search results with the metaindex of the node A, wherein the one additional metaindex merged to the metaindex of the node A includes an acquisition path, and sending the metaindex of the node A to the node X.Type: ApplicationFiled: July 15, 2008Publication date: June 18, 2009Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Fumiko Satoh, Toyotaro Suzumura, Issei Yoshida
-
Patent number: 7539683Abstract: Systems, methods, and computer products for information sharing using personalized index caching. Exemplary embodiments include a method including receiving a search query history from a node X in a node A, extracting characteristics of an index of the node A, searching the extracted characteristics, which include a file ID that is included in the index of the node A, adding metadata information to the index of the node A, in response to a determination that the node A includes at least one additional local metaindex, searching the at least one additional metaindex with the search query history from the node X in the node A, and merging search results with the metaindex of the node A, wherein the one additional metaindex merged to the metaindex of the node A includes an acquisition path, and sending the metaindex of the node A to the node X.Type: GrantFiled: July 15, 2008Date of Patent: May 26, 2009Assignee: International Business Machines CorporationInventors: Fumiko Satoh, Toyotaro Suzumura, Issei Yoshida
-
Publication number: 20090030892Abstract: A system of the present invention stores: a first index which designates lists of keywords contained in texts from identifications of the respective texts; a second index which designates lists of texts containing keywords from identifications of the respective keywords; and the number of texts containing the respective keywords. Then, upon receiving an input of a text search condition, the system calculates an estimation of search time by the first index and an estimation of search time by the second index, and determines which one of the first and second indexes makes a search faster. Then, by using the index which has been determined to make the search faster, the system searches for keywords which appear in texts satisfying the text search condition with higher frequency.Type: ApplicationFiled: March 26, 2008Publication date: January 29, 2009Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATIONInventors: Daisuke Takuma, Issei Yoshida, Yuta Tsuboi
-
Publication number: 20080319987Abstract: An entire document set is decomposed into a sum of subsets each having no common part. Next, a set of keywords appearing in each of the subsets divided in the aforementioned manner is categorized into groups on the basis of a remainder resulting from dividing a hash value of each of the keywords by a certain fixed integer value. Thereby, index files for the respective groups are created. Among the index files prepared for the respective subsets of the document in the aforementioned manner, ones each having the same group number are merged. Thereby, integrated index files corresponding to the respective individual group numbers are created. Such index files, however, exist as many as the number of group numbers, and have not yet become an index corresponding to the entire document set. In this respect, the index files existing as many as the number of group numbers are next merged into one, and thereby, an index file corresponding to the entire document set is created.Type: ApplicationFiled: June 3, 2008Publication date: December 25, 2008Inventors: Daisuke Takuma, Issei Yoshida
-
Patent number: 7440952Abstract: Systems, methods, and computer products for information sharing using personalized index caching. Exemplary embodiments include a method including receiving a search query history from a node X in a node A, extracting characteristics of an index of the node A, searching the extracted characteristics, which include a file ID that is included in the index of the node A, adding metadata information to the index of the node A, in response to a determination that the node A includes at least one additional local metaindex, searching the at least one additional metaindex with the search query history from the node X in the node A, and merging search results with the metaindex of the node A, wherein the one additional metaindex merged to the metaindex of the node A includes an acquisition path, and sending the metaindex of the node A to the node X.Type: GrantFiled: December 18, 2007Date of Patent: October 21, 2008Assignee: International Business Machines CorporationInventors: Fumiko Satoh, Toyotaro Suzumura, Issei Yoshida
-
Publication number: 20070136274Abstract: A system of the present invention stores: a first index which designates lists of keywords contained in texts from identifications of the respective texts; a second index which designates lists of texts containing keywords from identifications of the respective keywords; and the number of texts containing the respective keywords. Then, upon receiving an input of a text search condition, the system calculates an estimation of search time by the first index and an estimation of search time by the second index, and determines which one of the first and second indexes makes a search faster. Then, by using the index which has been determined to make the search faster, the system searches for keywords which appear in texts satisfying the text search condition with higher frequency.Type: ApplicationFiled: November 2, 2006Publication date: June 14, 2007Inventors: Daisuke Takuma, Issei Yoshida, Yuta Tsuboi
-
Publication number: 20070100686Abstract: Embodiments herein properly determine relevance between persons. Embodiments herein provide a system for evaluating relevance between persons, including relevance detecting means for detecting directed relevance of a person as a target to another person based on a text in which evaluations or feelings of the other person by another person is expressed; storing means for storing the detected relevance; evaluating means for evaluating, with respect to a set of a first person and second person to be evaluated, relevance between the first person and the second person based on a plurality of relevances including the stored relevances and reaching the second person from the first person via at least one further person; and outputting means for outputting the relevance evaluated.Type: ApplicationFiled: October 24, 2006Publication date: May 3, 2007Inventors: Akiko Murakami, Yuta Tsuboi, Hideo Watanabe, Issei Yoshida