发明名称 Generating caption based computer data
摘要 Correcting data by expediting identification of a part of the data, which needs to be corrected. In response to receiving an instruction, outputting the information, and identifying the part which needs to be corrected. In response to a received instruction, expediting correction of the part by correcting the part and outputting the information.
申请公布号 US9460065(B2) 申请公布日期 2016.10.04
申请号 US201213409712 申请日期 2012.03.01
申请人 INTERNATIONAL BUSINESS MACHINES CORPORATION 发明人 Miyamoto Kohtaroh;Shoji Midori
分类号 G06F17/00;G06F17/24 主分类号 G06F17/00
代理机构 Cantor Colburn LLPX 代理人 Cantor Colburn LLPX ;Chiu Steven
主权项 1. A computer system for assisting correcting data that has been created on a basis of an information unit, the computer system comprising: a first computer for outputting one or more of said information unit, and for identifying data which needs to be corrected in response to a received instruction, the data comprising a plurality of parts of data; and a plurality of client computers for outputting said information unit, and for correcting said plurality of parts of data which have been identified by the first computer in response to said received instruction; and an allocation determination unit, the allocation determination unit retrieving each of the one or more information units and determining an allocation of said one or more information units to the plurality of client computers, the determining comprising: determining a number of vacant client computers;determining a number of said one or more said information units have not been processed;selecting, for processing of one of the one or more of said information units, a vacant client computer responsive to determining that the number of said one or more said information units that have not been processed is greater than one, and the number of vacant client computers is one, wherein processing requirements of said one of the one or more of said information units matches processing requirements of the vacant client computer;selecting, for processing of one of the one or more of said information units, a vacant client computer responsive to determining that the number of said one or more said information units that have not been processed is one, and the number of vacant client computers is greater than one, the selecting comprising calculating a job optimum degree, the job optimum degree based on a weighted latency of each of the vacant client computers, wherein the weighted latency does not correspond linearly with the latency, and the processing is selected to be performed on the vacant client computer with a maximum job optimum degree; andselecting delaying the processing of the one or more of said information units by a vacant client computer responsive to determining that the number of said one or more said information units that have not been processed is one, and the number vacant client computers is one, the selecting delaying comprising calculating a job optimum degree, the job optimum degree based on a weighted latency of the vacant client computer, wherein the weighted latency does not correspond linearly with the latency, and selecting to prevent processing of the one of said one or more information units by the vacant client computer in response to determining that the job optimum degree exceeds a threshold, wherein said first computer distributes said information unit, one of said plurality of said parts of data which needs to be corrected, and one or more correcting instructions, to each of said plurality of client computers for parallel processing, each of said plurality of client computers automatically outputting said information unit which corresponds to said one of said plurality of parts of data, prior to correcting said plurality of parts of data.
地址 Armonk NY US