“…More recently Yang [39] obtained a break-even point, where precision equals recall, of .85 on the Reuters newswire collection. Other statistical and machine learning approaches to text categorization include: Maron [24], Crawford et al [10], Yang and Chute [40] Lewis and Gale [21], Apte et al [1], Lewis et al [23], Cohen and Singer [7,8], Hodges et al [16], Moulinier et al [27], Yang [38], Leung and Kan [19], Koller and Sahami [18], and Yang and Pedersen [41]. This paper describes text categorization experiments in the legal domain.…”