TY - GEN
T1 - The effect of query length on normalisation in information retrieval
AU - Cummins, Ronan
AU - O'Riordan, Colm
PY - 2010
Y1 - 2010
N2 - Document length normalisation is known to be a difficult problem in IR, as tuning is often needed to overcome the collection dependence problem known to affect many normalisation schemes. Furthermore, it has been shown in various studies that the most optimal level of normalisation to apply is correlated with query length. In this paper, we confirm this correlation and present experiments which investigates and explains the effect of query length on normalisation.
AB - Document length normalisation is known to be a difficult problem in IR, as tuning is often needed to overcome the collection dependence problem known to affect many normalisation schemes. Furthermore, it has been shown in various studies that the most optimal level of normalisation to apply is correlated with query length. In this paper, we confirm this correlation and present experiments which investigates and explains the effect of query length on normalisation.
UR - https://www.scopus.com/pages/publications/78650149426
U2 - 10.1007/978-3-642-17080-5_5
DO - 10.1007/978-3-642-17080-5_5
M3 - Conference Publication
AN - SCOPUS:78650149426
SN - 364217079X
SN - 9783642170799
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 26
EP - 32
BT - Artificial Intelligence and Cognitive Science - 20th Irish Conference, AICS 2009, Revised Selected Papers
T2 - 20th Annual Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2009
Y2 - 19 August 2009 through 21 August 2009
ER -