Learning subjective nouns using extraction pattern bootstrapping

Learning Subjective Nouns using Extraction Pattern Bootstrapping∗ Ellen Riloff School of Computing University of Utah Salt Lake City, UT 84112 riloff@cs.utah.edu Janyce Wiebe Department of Computer Science University of Pittsburgh Pittsburgh, PA 15260 wiebe@cs.pitt.edu Theresa Wilson Intelligent Systems Program University of Pittsburgh Pittsburgh, PA 15260 twilson@cs.pitt.edu Abstract We explore the idea of creating a subjectiv- ity classifier that uses lists of subjective nouns learned by bootstrapping algorithms. The goal of our research is to develop a system that can distinguish subjective sentences from ob- jective sentences. First, we use two bootstrap- ping algorithms that exploit extraction patterns to learn sets of subjective nouns. Then we train a Naive Bayes classifier using the subjec- tive nouns, discourse features, and subjectivity clues identified in prior research. The boot- strapping algorithms learned over 1000 subjec- tive nouns, and the subjectivity classifier per- formed well, achieving 77% recall with 81% precision. 1 Introduction Many natural language processing applications could benefit from being able to distinguish between factual and subjective information. Subjective remarks come in a variety of forms, including opinions, rants, allega- tions, accusations, suspicions, and speculation. Ideally, information extraction systems should be able to distin- guish between factual information (which should be ex- tracted) and non-factual information (which should be discarded or labeled as uncertain). Question answering systems should distinguish between factual and specula- tive answers. Multi-perspective question answering aims to present multiple answers to the user based upon specu- lation or opinions derived from different sources. Multi- ∗ This work was supported in part by the National Sci- ence Foundation under grants IIS-0208798 and IRI-9704240. The data preparation was performed in support of the North- east Regional Reseach Center (NRRC) which is sponsored by the Advanced Research and Development Activity (ARDA), a U.S. Government entity which sponsors and promotes research of import to the Intelligence Community which includes but is not limited to the CIA, DIA, NSA, NIMA, and NRO. document summarization systems need to summarize dif- ferent opinions and perspectives. Spam filtering systems must recognize rants and emotional tirades, among other things. In general, nearly any system that seeks to iden- tify information could benefit from being able to separate factual and subjective information. Subjective language has been previously studied in fields such as linguistics, literary theory, psychology, and content analysis. Some manually-developed knowledge resources exist, but there is no comprehensive dictionary of subjective language. Meta-Bootstrapping (Riloff and Jones, 1999) and Basilisk (Thelen and Riloff, 2002) are bootstrapping al- gorithms that use automatically generated extraction pat- terns to identify words belonging to a semantic cate- gory. We hypothesized that extraction patterns could also identify subjective words. For example, the pat- tern “expressed <direct object>” often extracts subjec- tive nouns, such as “concern”, “hope”, and “support”. Furthermore, these bootstrapping algorithms require only a handful of seed words and unannotated texts for train- ing; no annotated data is needed at all. In this paper, we use the Meta-Bootstrapping and Basilisk algorithms to learn lists of subjective nouns from a large collection of unannotated …

Content truncated to 500 words