SJTULTLAB: Chunk Based Method for Keyphrase Extraction

Proceedings of the 5th International Workshop on Semantic Evaluation, ACL 2010, pages 158–161, Uppsala, Sweden, 15-16 July 2010. c⃝2010 Association for Computational Linguistics SJTULTLAB: Chunk Based Method for Keyphrase Extraction Letian Wang Department of Computer Science & Engineering Shanghai Jiao Tong University Shanghai, China koh@sjtu.edu.cn Fang Li Department of Computer Science & Engineering Shanghai Jiao Tong University Shanghai, China fli@sjtu.edu.cn Abstract In this paper we present a chunk based keyphrase extraction method for scientific articles. Different from most previous sys- tems, supervised machine learning algo- rithms are not used in our system. Instead, document structure information is used to remove unimportant contents; Chunk ex- traction and filtering is used to reduce the quantity of candidates; Keywords are used to filter the candidates before generating final keyphrases. Our experimental results on test data show that the method works better than the baseline systems and is comparable with other known algorithms. 1 Introduction Keyphrases are sequences of words which cap- ture the main topics discussed in a document. Keyphrases are very useful in many natural lan- guage processing (NLP) applications such as doc- ument summarization, classification and cluster- ing. But it is an expensive and time-consuming job for users to tag keyphrases of a document. These needs motivate methods for automatic keyphrase extraction. Most existing algorithms for keyphrase extrac- tion treat this task as a supervised classifica- tion task. The KEA algorithm (Gordon et al., 1999) identifies candidate keyphrases using lex- ical methods, calculates feature values for each candidate, and uses a machine-learning algorithm to predict which candidates are good keyphrases. A domain-specific method (Frank et al., 1999) was proposed based on the Naive Bayes learn- ing scheme. Turney (Turney, 2000) treated a document as a set of phrases, which the learn- ing algorithm must learn to classify as positive or negative examples of keyphrases. Turney (Tur- ney, 2003) also presented enhancements to the KEA keyphrase extraction algorithm that are de- signed to increase the coherence of the extracted keyphrases. Nguyen and yen Kan (Nguyen and yen Kan, 2007) presented a keyphrase extraction algorithm for scientific publications. They also in- troduced two features that capture the positions of phrases and salient morphological phenom- ena. Wu and Agogino (Wu and Agogino, 2004) proposed an automated keyphrase extraction al- gorithm using a nondominated sorting multi- objective genetic algorithm. Kumar and Srinathan (Kumar and Srinathan, 2008) used n-gram filtra- tion technique and weight of words for keyphrase extraction from scientific articles. For this evaluation task, Kim and Kan (Kim and Kan, 2009) tackled two major issues in au- tomatic keyphrase extraction using scientific ar- ticles: candidate selection and feature engineer- ing. They also re-examined the existing features broadly used for the supervised approach. Different from previous systems, our system uses a chunk based method to extract keyphrases from scientific articles. Domain-specific informa- tion is used to find out useful parts in a document. The chunk based method is used to extract candi- dates of keyphrases in a document. Keywords of a document are used to …

Content truncated to 500 words