Abstract | ||
---|---|---|
In this paper, we propose two new methods target- ing NTCIR-4 QAC2. First, we use knowledge resem- bling "common sense" for question answering pur- poses. For example, the length of a runway in an air- port must be a few kilometers, but a few centimeters. In practice, we use specific types of information la- tent in document collections to verify the correctness of each answer candidate. Second, we use the utility maximization principle to determine the appropriate number of answers for a list question. We estimate the expected value of the evaluation score, on the basis of the probability scores for multiple answer candidates. We show the effectiveness of our methods by means of experiments. |
Year | Venue | Field |
---|---|---|
2004 | NTCIR | Common sense,Question answering,Information retrieval,Correctness,Utility maximization,Expected value,Runway,Mathematics |
DocType | Citations | PageRank |
Conference | 11 | 1.04 |
References | Authors | |
12 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tomoyosi Akiba | 1 | 176 | 29.08 |
Katunobu Itou | 2 | 319 | 44.36 |
Atsushi Fujii | 3 | 486 | 59.25 |