Bio
Dora Alexopoulou is a linguist, interested in natural language syntax, as a crucial manifestation of the unique human cognitive capacity for language, and the way it shapes language learning. She is an Assistant Professor in Language Acquisition and Language Typology and PI of the EF Lab for Applied Language Learning. Her background is in Greek philology (BA, Athens), Natural Language Processing (MSc, Edinburgh) and Linguistics (PhD, Edinburgh).
Theodora’s research looks at variation and diversity across languages and the way in which the linguistic distance between a learner’s mother tongue and their second/additional language influences the way they process and interpret their additional language, how it impacts on their learning, even predicting proficiency scores in exams. Her goal is to develop learning methods and tools that can address learning challenges and improve outcomes. I approach these questions through analysing data from online language learning platforms which can provide us with large and rich data from learners from a variety of educational, cultural and socio-economic contexts across the globe. More recently, I have started considering the neural correlates of native language influence on second language learners.
Research interests
Data science can revolutionalise research in language learning with direct impact on language education to help improve UK’s language skills. Language and learning data from online language learning platforms create unprecedented research opportunities. Such data allow us to consider how linguistic (e.g. mother tongue), cognitive (e.g. age, motivation) and socio-cultural characteristics of individual learners interact with the characteristics of different educational settings and socio-cultural contexts to impact on the learning outcomes.
How can we create data sets for research from the millions of daily interactions on language learning platforms? How can we extend current Natural Language Technology to extract patterns that better capture variation across learners and contexts? How can we combine insights from linguists, cognitive scientists and educators who understand the nature of the different variables and data scientists who understand the modelling challenges? Ηow can we bring insights from our human understanding of the learning process and individual learner needs to AI for education?