Learning To Simplify Text One Sentence at a Time
The complexity and readability of text can vary widely. Compare, for example, a research article and a popular media article on the same topic. The articles may discuss the same material, but their usefulness for the reader can vary widely due to the accessibility of the language and structure of the articles.
In this talk I will examine the problem of text simplification which aims to develop models to automatically reduce the reading complexity of text. Advances in automated text simplification have a broad set of applications ranging from medicine to second language learners. Motivated by both corpus analysis and human experiments, I will introduce a number of recent simplification models and techniques that leverage a variety of data sources to automatically learn simplifications models.
David Kauchak is an Assistant Professor at Middlebury College in the Computer Science Department. His research interests are at the intersection of machine learning and natural language processing. He has previously taught at Pomona College and Stanford University and has spent time in industry at companies ranging from tech startups to larger institutes including Google and PARC.