Classifying easy-to-read texts without parsing
2014 (English)In: Proceedings of the 3rd Workshop on Predicting and Improving Text Readability for Target Reader Populations (PITR), 2014, 114-122 p.Conference paper (Refereed)
Document classification using automated linguistic analysis and machine learning (ML) has been shown to be a viable road forward for readability assessment. The best models can be trained to decide if a text is easy to read or not with very high accuracy, e.g. a model using 117 parameters from shallow, lexical, morphological and syntactic analyses achieves 98,9% accuracy. In this paper we compare models created by parameter optimization over subsets of that total model to find out to which extent different high-performing models tend to consist of the same parameters and if it is possible to find models that only use features not requiring parsing. We used a genetic algorithm to systematically optimize parameter sets of fixed sizes using accuracy of a Support Vector Machine classi- fier as fitness function. Our results show that it is possible to find models almost as good as the currently best models while omitting parsing based features.
Place, publisher, year, edition, pages
2014. 114-122 p.
Readability, Readability Assessment, Genetic optimization, Machine Learning, Support Vector Machine
Language Technology (Computational Linguistics)
IdentifiersURN: urn:nbn:se:liu:diva-117547ISBN: 978-1-937284-91-6OAI: oai:DiVA.org:liu-117547DiVA: diva2:809460
14th Conference of the European Chapter of the Association for Computational Linguistics