@@ -7,15 +7,15 @@ Each line contains ID of QA pair and an answer of QA.
77
88##testset.txt
99Each line is tab-separated.
10- First column is user's query.
11- The 2nd,3th and 4th columns are the QA pair IDs in relevance level A, B and C, respectively.
12- The IDscorresponds to ones in qas/questions_Amagasaki.txt and qas/answers_in_Amagasaki.txt.
10+ The first column is user's query.
11+ The 2nd,3rd and 4th columns are the QA pair IDs in relevance level A, B and C, respectively.
12+ The IDscorrespond to ones in qas/questions_Amagasaki.txt and qas/answers_in_Amagasaki.txt.
1313The description about relevance levels is below.
1414```
1515A, Contain correct information.
1616B, Contain relevant information.
17- C, The topic is same as a query, butdo not contain relevant information.
17+ C, The topic is same as a query, butdoes not contain relevant information.
1818```
1919
2020##testset_segmentation.txt
21- testset.txtafter wordsegmentation by[ ` Juman++ ` ] ( http://nlp.ist.i.kyoto-u.ac.jp/EN/index.php?JUMAN++ )
21+ testset.txtwith wordsegmentations by[ ` Juman++ ` ] ( http://nlp.ist.i.kyoto-u.ac.jp/EN/index.php?JUMAN++ )