A Tagging Approach to Identify Complex Constituents for Text Simplification

Proceedings of Recent Advances in Natural Language Processing, pages 221–229, Hissar, Bulgaria, 7-13 September 2013. A Tagging Approach to Identify Complex Constituents for Text Simplification Iustin Dornescu Richard Evans Research Institute in Information and Language Processing University of Wolverhampton United Kingdom {I.Dornescu2, R.J.Evans, C.Orasan}@wlv.ac.uk Constantin Or˘asan Abstract The occurrence of syntactic phenomena such as coordination and subordination is characteristic of long, complex sentences. Text simplification systems need to detect and categorise constituents in order to generate simpler sentences. These con- stituents are typically bounded or linked by signs of syntactic complexity, which in- clude conjunctions, complementisers, wh- words, and punctuation marks. This paper proposes a supervised tagging approach to classify these signs in accordance with their linking and bounding functions. The performance of the approach is evaluated both intrinsically, using an annotated cor- pus covering three different genres, and extrinsically, by evaluating the impact of classification errors on an automatic text simplification system. The results are encouraging. 1 Introduction This paper presents an automatic method to de- termine the specific coordinating and bounding functions of several reliable signs of syntactic complexity in natural language. This method can be useful for automatic text simplification. The syntactic complexity of input text can be reduced by the application of rules triggered by patterns expressed in terms of the parts of speech of words and the syntactic linking and bounding functions of signs of syntactic complexity occurring within it (Evans, 2011). Previous work indicates that syntactic simplification can improve text accessi- bility (Just et al., 1996) and the reliability of NLP applications such as information extraction (Agar- wal and Boggess, 1992; Rindflesch et al., 2000), machine translation (Gerber and Hovy, 1998), and syntactic parsing (Tomita, 1985; McDonald and Nivre, 2011). The research described in the current paper is part of the FIRST project1 which aims to automatically convert documents into a more accessible form for people with autistic spectrum disorders (ASD). Many of the decisions taken in the research presented in this paper were informed by the psycholinguistic experiments carried out in the FIRST project and summarised in Martos et al. (2013). The remainder of this paper is structured as fol- lows. Section 2 provides background information about the context of this work, Section 3 presents the annotation scheme, Section 4 describes the approach and the main objectives of this study. The results and the main findings are presented in Section 5. Section 6 provides an overview of previous related work. In Section 7, conclusions are drawn. 2 Syntactic Simplification in the FIRST Project Research carried out in the FIRST project and investigation of related work revealed that certain types of syntactic complexity adversely affect the reading comprehension of people with ASD (Martos et al., 2013). This section presents a brief overview of the context in which this research is carried out. It builds on the approach proposed by Evans (2011) who presented a rule- based method to simplify sentences containing coordinated constituents to facilitate information extraction. In that work, punctuation marks and conjunctions …

Content truncated to 500 words