2006). We present our recent work on the task of
the NER extraction, we also employ phrase analy-
QA, wherein systems aim at determining if a text
sis based on our phrase utility extraction method
returned by a search engine contains the correct
using Standford dependency parser ((Klein and
answer to the question posed by the user. The ma-
Manning, 2003)). We can categorize entities up
jor categories of information extraction produced
to 6 coarse and 50 fine categories to match them
by our QA system characterizes features for our
with the NER types from QC module.
TE model based on analysis of q/a pairs. Here we
Phrase Identification(PI): Our PI module un-
give brief descriptions of only the major modules
dertakes basic syntactic analysis (shallow pars-
of our QA due to space limitations.
ing) and establishes simple, un-embedded linguis-
Bạn đang xem 2006) - TÀI LIỆU BÁO CÁO KHOA HỌC: "A GRAPH-BASED SEMI-SUPERVISED LEARNING FOR QUESTION-ANSWERING" DOC