会議名: 情報知識学会 第11回(2003年度)研究報告会
This work is that tring to extract the relationships of case and effect between method and effect to notice specially essential description of invention's method and effect in patent documents by text process using syntactic rule. we also extract other relationships based on getting relationships. Then, we mention thinking machine model using extracted relationships cause and effect. Notice this work use distributed patent corpus of NTCIR-3 in 2002.