发明名称 |
ANALYSIS OF EVALUATIONS FROM INTERNET MEDIA |
摘要 |
Embodiments relate to supporting analysis of evaluations found on the internet. An aspect includes acquiring evaluation information comprising evaluations of each a first plurality of evaluation targets made by any of a plurality of evaluators. Another aspect includes generating consensus information regarding an evaluation of a predetermined evaluation target by determining a coincidence between a dispersion degree of evaluations with respect to the predetermined evaluation target and a central value of dispersion degrees of evaluations with respect to individual ones of the first plurality of evaluation targets. Another aspect includes generating distinctiveness information regarding the evaluations made by a predetermined evaluator by determining coincidence between a dispersion degree of evaluations made by the predetermined evaluator and a central value of dispersion degrees of evaluations made by individual ones of the plurality of evaluators. Yet another aspect includes outputting the consensus information and the distinctiveness information. |
申请公布号 |
US2015170046(A1) |
申请公布日期 |
2015.06.18 |
申请号 |
US201414554487 |
申请日期 |
2014.11.26 |
申请人 |
International Business Machines Corporation |
发明人 |
Sugiyama Yuji |
分类号 |
G06N5/04;H04L29/08 |
主分类号 |
G06N5/04 |
代理机构 |
|
代理人 |
|
主权项 |
1. A method for supporting analysis of evaluations found on the internet, comprising:
acquiring evaluation information comprising evaluations of each a first plurality of evaluation targets made by any of a plurality of evaluators evaluating any of the first plurality of evaluation targets, and evaluations made by each of the plurality of evaluators with respect to evaluation targets out of a second plurality of evaluation targets evaluated by any of the plurality of evaluators, wherein a number of the second plurality of evaluation targets is greater than or equal to a number of the first plurality of evaluation targets; generating consensus information regarding an evaluation of a predetermined evaluation target by determining a coincidence between a dispersion degree of evaluations with respect to the predetermined evaluation target and a central value of dispersion degrees of evaluations with respect to individual ones of the first plurality of evaluation targets based on the evaluation information; generating distinctiveness information regarding the evaluations made by a predetermined evaluator by determining coincidence between a dispersion degree of evaluations made by the predetermined evaluator and a central value of dispersion degrees of evaluations made by individual ones of the plurality of evaluators based on the evaluation information; and outputting the consensus information and the distinctiveness information. |
地址 |
Armonk NY US |