Title:
METHOD FOR EXTRACTING MAIN FEATURE FROM VOICE OF CUSTOMER DATA AND METHOD FOR CLASSIFYING DATA TYPE BY USING SAME
Document Type and Number:
WIPO Patent Application WO/2019/189951
Kind Code:
A1
Abstract:
The present invention relates to a method for processing voice of customer (hereinafter, referred to as "VOC") data and, more specifically, to a method for extracting a main feature from VOC data on the basis of a lexico semantic pattern (hereinafter, referred to as "LSP") and classifying the type of newly inputted VOC data by using the extracted main feature. To this end, the present invention provides a method for classifying a VOC data type, the method allowing a word embedding performance result and previously constructed LSP information to be used on the basis of a semantic feature pattern so as to extend a semantic feature such that the quality of syntax analysis is improved, and enabling a syntax structure and semantic analysis and a type classification, which have not been conventionally performed, so as to help in decision making by allowing VOC data to be easily understood.
Inventors:
KIM MOON JONG (KR)
PARK TAE SOO (KR)
MOON YOUNG JU (KR)
LEE CHAN HYEONG (KR)
JANG JUNG HOON (KR)
PARK TAE SOO (KR)
MOON YOUNG JU (KR)
LEE CHAN HYEONG (KR)
JANG JUNG HOON (KR)
Application Number:
PCT/KR2018/003611
Publication Date:
October 03, 2019
Filing Date:
March 27, 2018
Export Citation:
Assignee:
WISENUT INC (KR)
International Classes:
G06Q30/02; G06F40/00; G10L15/08
Foreign References:
KR20170088095A | 2017-08-01 | |||
KR101339103B1 | 2013-12-09 | |||
KR20130022075A | 2013-03-06 | |||
JP2003162524A | 2003-06-06 |
Other References:
KIM MOONJONG, ET AL.: "VOC Summarization and Classification based on Sentence Understanding", THE KOREAN INSTITUTE OF INFORMATION SCIENTISTS AND ENGINEERS, KIISE TRANSACTIONS ON COMPUTING PRACTICES, vol. 22, no. 1, January 2016 (2016-01-01), pages 50 - 55
Attorney, Agent or Firm:
NURY PATENT LAW FIRM (KR)
Download PDF: