As a part of Automatic Functional Language Extraction (AFLEX) project at ISU, I am working on exploring ways to extraction different kinds of information from scientific articles (e.g., what is the study design? what are the methods used? what are the main results? etc.). I work with Raghul Ethiraj (Graduate Research Assistant), and the AFLEX project team in this project. Within the team, I collaborate actively with Dr Zhu Zhang, Dr Elena Cotos, Dr Stephen Gilbert, Dr Carol Chapelle and Dr Vijay Kalivarapu.
Research Writing Tutor (RWT) project in ISU deals with providing students online feedback on writing research articles, comparing them with articles written in their own disciplines. I also work with the RWT team, to develop machine learning methods that can be used to classify sentences into different rhetorical categories. I work with Dr Elena Cotos and her team in the RWT team.
In this strand of research, I am interested in analyzing the writing of non-native English speakers for two aspects: a) automatically scoring their proficiency, b) understanding if we can detect any patterns of their native language (L1) in their English writing. For the second part, I worked closely with Dr Magdalena Wolska, Universtiy of Tuebingen, Germany in the past. I now work with Liberato Silva-Dos-Santos and Rosalie Hirch (Graduate Students at ISU) on this.
Along with Ziwei Zhou (graduate student at ISU), I recently started working on analysing non-native speech for scoring and accent identification (More on this in future!!)
Automatic Readability Assessment refers to the task of predicting the reading difficulty for a given text, for a target reader. I worked on developing machine learning models for readability assessment, and on integrating text and user characteristics into a single model. As a part of understanding text-user interaction, I was involved in a eye-tracking study in the past (CL4LC paper, 2016), and recently finished a larger reading study with over 100 participants along with Ivana Lucic (Graduate student at ISU) and Stephanie Fuccio (ex-student at ISU). More details on the findings from this study will come soon!
This is an inter-disciplinary collaboration with Educational Psychologists at the University of Tuebingen, where we look at the linguistic complexity of Geography textbooks in some German schools, at different grade levels and school types. More details on this project can be found on the project homepage. Our paper on this project is published in the Journal of Educational Psychology: Berendes, K., Vajjala, S., Meurers, D., Bryant, D., Wagner, W., Chinkina, M., & Trautwein, U. (2017). Reading Demands in Secondary School: Does the Linguistic Complexity of Textbooks Increase With Grade Level and the Academic Orientation of the School Track?. Journal of Educational Psychology. Advance online publication. Pre-print. I gave a presentation about this at a Brownbag session in ISU. Slides are here (November 2017)
Sowmya Vajjala. 2017. "Automated assessment of non-native learner essays: Investigating the role of linguistic features" International Journal of Artificial Intelligence in Education, 2017. ( pre-print. code. article on publisher website.
Towards grounding computational lingistic approaches to readability: Modeling reader-text interaction for easy and difficult texts. Proceedings of Computational Linguistics for Linguistic Complexity Workshop in COLING 2016.
Authors: Sowmya Vajjala, Detmar Meurers, Katharina Scheiter, Alexander Eitel. 2016. (article).Recent Conference Presentations (abstracts)
In 2016, I participated in American Association of Corpus Linguistics conference (AACL 2016) and presented our work on two research projects.
a) Visualizing the relationship between language proficiency and native language in a corpus of learner essays (with Magdalena Wolksa). Abstract. Talk slides.
a) Computational analysis of methods sections in a corpus of scientific articles (with Elena Cotos, Carol Chapelle and Hyunwoo Kim). Abstract. Poster.