TY - JOUR
T1 - Experiment on combining sources of evidence for passage retrieval
AU - Gelbukh, Alexander
AU - Kang, Nam O.
AU - Han, Sangyong
PY - 2005
Y1 - 2005
N2 - Passage retrieval consists in identifying short but informative runs of a long text, given a specific user query. We discuss the sources of evidence that help choosing likely high-quality passages, such as relevance to the user query and self-containedness. These measures are different from the traditional information retrieval procedure due to the use of the context of the passage.
AB - Passage retrieval consists in identifying short but informative runs of a long text, given a specific user query. We discuss the sources of evidence that help choosing likely high-quality passages, such as relevance to the user query and self-containedness. These measures are different from the traditional information retrieval procedure due to the use of the context of the passage.
UR - http://www.scopus.com/inward/record.url?scp=24344508234&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-30586-6_69
DO - 10.1007/978-3-540-30586-6_69
M3 - Artículo de la conferencia
AN - SCOPUS:24344508234
SN - 0302-9743
VL - 3406
SP - 628
EP - 631
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
T2 - 6th International Conference, CICLing 2005
Y2 - 13 February 2005 through 19 February 2005
ER -