Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
An extracting apparatus, an extraction method, and a program
Document Type and Number:
Japanese Patent JP5955817
Kind Code:
B2
Abstract:
PROBLEM TO BE SOLVED: To allow accurate extraction of relevant words from stream data.SOLUTION: An extraction device sums up, for each of preliminarily determined words, an appearance frequency with which the word appears in input stream data, at intervals of a prescribed time and stores the summed-up appearance frequency of the word in every prescribed time and records the appearance frequency summed up until the present time, of the word. The extraction device detects burst of words in the prescribed time on the basis of summed-up appearance frequencies in every prescribed time of respective words and recorded appearance frequencies summed up until the present time, of the words and stores, for each of words of which the burst has been detected, burst feature quantity information including the appearance frequency in the prescribed time. The extraction device accepts a period and a reference word as an extraction condition and normalizes appearance frequencies of respective words in burst feature quantity information in the period and extracts, as relevant words, words of which the appearance frequencies after normalization are similar to the appearance frequency after normalization of the reference word.

Inventors:
Ra Min
Han Seong-Ke
Application Number:
JP2013154872A
Publication Date:
July 20, 2016
Filing Date:
July 25, 2013
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
Nippon Telegraph and Telephone Corporation
International Classes:
G06F17/30
Domestic Patent References:
JP2008250975A
JP2000194745A
Foreign References:
US20070121674
Other References:
久野雄一郎、澤勢一史、延原肇,Twitterにおける極大部分文字列の反復度および時系列相関を用いた類似単語判定,電子情報通信学会技術研究報告,日本,一般社団法人電子情報通信学会,2013年 2月28日,Vol.112,No.465,pp. 21-26
風間一洋、鳥海不二夫、榊剛史、篠田孝祐、栗原 聡、野田五十樹,東日本大震災時のTwitterデータを用いた単語間の関係の時系列変化の分析,2012年度人工知能学会全国大会(第26回)論文集 [CD-ROM],日本,社団法人人工知能学会,2012年 6月12日,pp. 1-4
Attorney, Agent or Firm:
Sakai International Patent Office