-
- いいね！0
Naoki Fukuta(福田直樹)
@fukutax
Q. In general, we add some ideas for applying the algorithm rather than just applying it straightforward. #deim2010 A8-1
2010-03-01 17:07:32
Naoki Fukuta(福田直樹)
@fukutax
Q. Do you consider any approaches about an automation mechanism for it? A. Yes, but there are other issues such as ranking. #deim2010 A8-1
2010-03-01 17:04:47
Naoki Fukuta(福田直樹)
@fukutax
Q. In your approach does the user choose scoring functions manually? A. Yes. Also we evaluated how the choice affects. #deim2010 A8-1
2010-03-01 17:02:37
Naoki Fukuta(福田直樹)
@fukutax
A. Our approach is another approach and intentionally did not consider contexts of appearing the words. #deim2010 A8-1
2010-03-01 17:01:13
Naoki Fukuta(福田直樹)
@fukutax
Q. Do you know a well-known work presented in ACL2002? That is used similar approach and consider contexts in the page. #deim2010 A8-1
2010-03-01 16:59:38
Naoki Fukuta(福田直樹)
@fukutax
Q. Did you consider the context of the words for generating the answer? #deim2010 A8-1
2010-03-01 16:58:30
Ohue M/大上雅史
@tonets
Q:次元数が２０ぐらいで次元縮約する意味がわからないが A:次元縮約を行うことで的中率が上がるかもしれないという試み． B8-1 #deim2010
2010-03-01 16:57:17
Naoki Fukuta(福田直樹)
@fukutax
Automatic construction of QA-type search systems based on machine learning #deim2010 A8-1
2010-03-01 16:43:30