The use of bigrams to enhance text categorization

By Chade-Meng Tan, Yuan-Fang Wang, Chansik Lee. Information Processing & Management, 2002

329 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

N-gram-Based Low-Dimensional Representation for Document Classification (2014, background): “A collection of words cannot capture phrases or multi-word expressions, while n-grams have shown to be helpful features in several natural language processing tasks (Tan et al., 2002; Lin & Wu, 2009; Wang & Manning, 2012).”
What are developers talking about? An analysis of topics and trends in Stack Overflow (2014, methodology): “Since 2-grams (equivalently, bi-grams) have been shown to increase the quality of text analysis (Tan et al. 2002), we use both uni-grams and bi-grams in the creation of the topics.”
Twitter Sentiment Analysis (2015, unknown): “Tan et al. reported that use of bigrams on Yahoo-Science dataset (Tan et al., 2002) allowed to improve the performance of text classiﬁcation using Naive Bayes classiﬁer from 65% to 70% break-even point, however, on Reuters-21578 dataset the increase of accuracy was not signiﬁcant.”
Controlled user evaluations of information visualization interfaces for text retrieval: Literature review and meta-analysis (2008, unknown): “However, good classification results can be achieved, as demonstrated by the many successful applications of ME to natural language processing (NLP) and IR problems reported in the literature, such as part-of-speech (POS) tagging (Ratnaparkhi, 1996a), noun-phrase extraction (Feng & Croft, 2001), doc”