SEERLAB: A System for Extracting Keyphrases from Scholarly Documents

Proceedings of the 5th International Workshop on Semantic Evaluation, ACL 2010, pages 182–185, Uppsala, Sweden, 15-16 July 2010. c⃝2010 Association for Computational Linguistics SEERLAB: A System for Extracting Keyphrases from Scholarly Documents Pucktada Treeratpituk1 Pradeep Teregowda2 Jian Huang1 C. Lee Giles1,2 1 Information Sciences and Technology 2 Computer Science and Engineering Pennsylvania State University, University Park, PA, USA Abstract We describe the SEERLAB system that participated in the SemEval 2010’s Keyphrase Extraction Task. SEERLAB utilizes the DBLP corpus for generating a set of candidate keyphrases from a document. Random Forest, a supervised ensemble classifier, is then used to select the top keyphrases from the candidate set. SEERLAB achieved a 0.24 F-score in generating the top 15 keyphrases, which places it sixth among 19 participating sys- tems. Additionally, SEERLAB performed particularly well in generating the top 5 keyphrases with an F-score that ranked third. 1 Introduction Keyphrases are phrases that represent the impor- tant topics of a document. There are two types of keyphrases associated with scholarly publications: author-assigned ones and reader-assigned ones. In the Keyphrase Extraction Task (Kim et al., 2010), each system receives two set of scientific papers from the ACM digital library; a training set and a testing set. The author-assigned keyphrases and reader-assigned keyphrases are given for each pa- per in the training set. The objective is to produce the keyphrases for each article in the testing set. This paper is organized as followed. First, We describe our keyphrase extraction system, SEER- LAB. We then discuss its performance in SemEval 2010. Lastly, we analyze the effectiveness of each feature used by SEERLAB, and provide a sum- mary of our findings. 2 System Description SEERLAB consists of three main components: a section parser, a candidate keyphrase extractor, and a keyphrase ranker. To generate keyphrases for a paper, the section parser first segments the document into pre-defined generic section types. Secondly, the candidate keyphrase extractor gen- erates a list of candidate phrases based on the doc- ument content. Then, the keyphrase ranker ranks each candidate according to the likelihood that it is a keyphrase. The top candidates are selected as keyphrases of the paper. 2.1 Section Parser The goal of the section parser is to parse each doc- ument into the same set of pre-defined sections. However, segmenting a scientific article into pre- defined section types is not trivial. While schol- arly publications generally contains similar sec- tions (such as Abstract and Conclusion), a sec- tion’s exact header description and the order in which it appears can vary from document to docu- ment. For example, the “Related Work” section is sometimes referred to as “Previous Research” or “Previous Work.” Also, while the “Related Work” section often appears right after the introduction, it could also appear near the end of a paper. (Nguyen and Kan, 2007) had success in us- ing a maximum entropy (ME) classifier to clas- sify sections into 14 generic section types includ- ing those such as Motivation, Acknowledgement, References. However, their approach requires an- notated training …

Content truncated to 500 words