Revisiting Document Length Hypotheses NTCIR-4 CLIR and Patent Experiments at Patolis
Sumio Fujita · NTCIR · 2004
This paper describes Patolis NTCIR-4 experiments of CLIR and Patent tasks, focusing on comparative studies of two test-collections and two retrieval strategies in view of document length hypotheses. TF*IDF outperformed language modeling approach in the CLIR task while two approaches performed similarly in the Patent task. We assumed two different document length hypotheses behind two collections. Some other task specific techniques are evaluated and reported.