A research group at the University of Tsukuba has collected a dataset of eye movements for more than 41 words when native Japanese speakers read English sentences, and has constructed and released it as the Tsukuba Eye-tracking Corpus (TECO).

 It has become clear that eye movement data during reading comprehension is useful for clarifying the process by which humans process and understand written language. For this reason, in recent years, efforts have been made to systematically organize large-scale eye movement data collected by a method called eye tracking during reading comprehension in one's native language or a foreign language and to make them publicly available as eye-tracking corpora. However, data on English language learners has mainly been collected in Europe and surrounding areas, and data from Asia has been lacking.

 The research group recruited 41 native Japanese-speaking undergraduate and graduate students who were English learners and collected eye movement data for approximately 2 words per participant (totaling more than 1 words) as they read English texts equivalent to the Eiken Grade Pre-1 to Grade Pre-1 levels.

 This dataset includes a total of nine indicators for each word in the text: the total number of fixations, the total duration of fixations, the duration of the first fixation, the duration of fixations before the next word is read, the duration of fixations before the next word is read including backreading, whether or not there was skipping, whether or not there was backreading from the subsequent word, whether or not there was a second fixation, and whether or not there was a second read. Analysis of these confirmed that the reliability and validity of the collected eye movement data was maintained.

 The collected data has been made publicly available on an open science platform as the Tsukuba Eye-tracking Corpus (TECO), and it is expected that it will be used in research on text comprehension and language acquisition. It is also expected to have a ripple effect on a wide range of academic fields, including natural language processing and AI research.

Paper information:[Research Methods in Applied Linguistics] TECO: An Eye-Tracking Corpus of Japanese L2 English Learnersʼ Text Reading

University of Tsukuba

Continue to challenge interdisciplinary fusion and internationalization, and develop human resources with intelligence and humanity

University of Tsukuba is the core of Tsukuba Science City, an international city with a high academic and cultural scent.The current education system is 9 groups, 23 classes, and it is possible to take specialized introductory courses from all fields, aiming to have creative intelligence and rich humanity.Solving global issues with soul genius […]

University Journal Online Editorial Department

This is the online editorial department of the university journal.
Articles are written by editorial staff who have a high level of knowledge and interest in universities and education.