Abstract: Map label translation implementations described herein transliterate and translate map labels in a first language, even those written in a non-Latin script, into a Latin script and into a second language. In some implementations a translation candidate generator process generates possible translation candidate sequences for each n-gram in the map label, for example, an entity's name in Japanese and possibly in native Japanese (non-Latin, non-Roman) script. A candidate sequence selector selects a number of top possible translation candidate sequences. A ranking feature extraction process is used to rank the selected top number of top possible translation candidate sequences by using a trained probabilistic classifier and geospatial and linguistic context information as ranking features. A post ranker then re-ranks the selected ranked translation candidates depending on neighboring proximity information around the location of the entity and outputs the best map label translation in the second language.

Abstract: Map label translation implementations described herein transliterate and translate map labels in a first language, even those written in a non-Latin script, into a Latin script and into a second language. In some implementations a translation candidate generator process generates possible translation candidate sequences for each n-gram in the map label, for example, an entity's name in Japanese and possibly in native Japanese (non-Latin, non-Roman) script. A candidate sequence selector selects a number of top possible translation candidate sequences. A ranking feature extraction process is used to rank the selected top number of top possible translation candidate sequences by using a trained probabilistic classifier and geospatial and linguistic context information as ranking features. A post ranker then re-ranks the selected ranked translation candidates depending on neighboring proximity information around the location of the entity and outputs the best map label translation in the second language.

Abstract: A method of driving an electrophoretic display device includes changing the gradation level of image data on the basis of correction data corresponding to the gradation level, converting image data with the changed gradation level to a dithering pattern, in which the first color and the second color are combined, corresponding to the changed gradation level for each predetermined region of image data, and driving the electrophoretic particles of the first color and the electrophoretic particles of the second color on the basis of image data converted to the dithering pattern for the plurality of pixels in the display section.

Abstract: A plurality of query responses is obtained based on a query. A corresponding selectable thumbnail map is determined for each query response, each corresponding selectable thumbnail map including a graphical geographic focus entity that is associated with the corresponding query response. A display of the query responses with the corresponding selectable thumbnail maps each displayed adjacent to its corresponding query response is initiated, in response to the query. A request for an expanded map is received based on a selection action associated with one of the corresponding selectable thumbnail maps. A display of the expanded map that includes a selectable link associated with the corresponding query response that is associated with the corresponding selectable thumbnail map that represents a compressed view of the expanded map is initiated, the expanded map displayed in a common display area with the display of the query responses with the corresponding selectable thumbnail maps.

Abstract: A file search system includes : a file analyze device including: a character string identifying unit; a first calculating unit; a second calculating unit; an analyzing unit; an output unit; and a file search device including a first obtaining unit; a second obtaining unit; an index storage unit; a file identifying unit; a ranking unit; and a notifying unit.

Abstract: A method of driving an electrophoretic display device includes changing the gradation level of image data on the basis of correction data corresponding to the gradation level, converting image data with the changed gradation level to a dithering pattern, in which the first color and the second color are combined, corresponding to the changed gradation level for each predetermined region of image data, and driving the electrophoretic particles of the first color and the electrophoretic particles of the second color on the basis of image data converted to the dithering pattern for the plurality of pixels in the display section.

Abstract: A device composed of structures and access methods for providing search functionalities is proposed. The structure includes a presence/absence information unit that stores bit-level items grouped in blocks. Other structures included are, a character string list that is used to store keys composed of character strings and a file specification unit that uniquely identify files from multiple files. Each element of the 2 former structures has a uniquely assigned identification number. The presence/absence information items associated with each file are ordered according to the identification numbers stated previously. The elements of the former 2 structures are many-to-many associated through the presence/absence information items, using a combination of identification numbers. An obtaining unit for search keys is provided, together with the output unit that produces file identifiers specified by the file specification unit.

Abstract: A file search system includes : a file analyze device including: a character string identifying unit; a first calculating unit; a second calculating unit; an analyzing unit; an output unit; and a file search device including a first obtaining unit; a second obtaining unit; an index storage unit; a file identifying unit; a ranking unit; and a notifying unit.

Abstract: A device composed of structures and access methods for providing search functionalities is proposed. The structure includes a presence/absence information unit that stores bit-level items grouped in blocks. Other structures included are, a character string list that is used to store keys composed of character strings and a file specification unit that uniquely identify files from multiple files. Each element of the 2 former structures has a uniquely assigned identification number. The presence/absence information items associated with each file are ordered according to the identification numbers stated previously. The elements of the former 2 structures are many-to-many associated through the presence/absence information items, using a combination of identification numbers. An obtaining unit for search keys is provided, together with the output unit that produces file identifiers specified by the file specification unit.