Pro-TEXT: an Annotated Corpus of Keystroke Logs

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 1732–1739 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 1732 Pro-TEXT: an Annotated Corpus of Keystroke Logs Aleksandra Mileti´c1, Christophe Benzitoun2, Georgeta Cislaru3, Santiago Herrera-Yanez1 1 Clesthia (Paris 3 University), 2 ATILF (CNRS & Lorraine University), 3 MoDyCo (CNRS & Paris Nanterre University) France {aleksandra.haddad, georgeta.cislaru, santiago.herrera-yanez}@sorbonne-nouvelle.fr, christophe.benzitoun@atilf.fr Abstract Pro-TEXT is a corpus of keystroke logs written in French. Keystroke logs are recordings of the writing process executed through a keyboard, which keep track of all actions taken by the writer (character additions, deletions, substitutions). As such, the Pro-TEXT corpus offers new insights into text genesis and underlying cognitive processes from the production perspective. A subset of the corpus is linguistically annotated with parts of speech, lemmas and syntactic dependencies, making it suitable for the study of interactions between linguistic and behavioural aspects of the writing process. The full corpus contains 202K tokens, while the annotated portion is currently 30K tokens large. The annotated content is progressively being made available in a database-like format, and the work on an HTML-based visualisation tool is currently under way. To the best of our knowledge, Pro-TEXT is the first corpus of its kind in French. Keywords: annotated corpus, keystroke logs, annotation methodology, writing process, French 1. Introduction This paper reports on annotation efforts on Pro-TEXT, a corpus based on keystroke logs written in French. Keystroke logs are recordings of the writing process executed on a keyboard and captured through dedicated software (Leijten and Van Waes, 2006; Str¨omqvist and Malmsten, 1998; Carl, 2012). These recordings keep track of all actions taken by the writer during the writ- ing process (character additions and deletions, mouse movements, copy-paste substitutions, etc.), making them well-suited for data-based studies on the dynam- ics of the writing process and the underlying cognitive mechanisms. This is illustrated by the wide variety of research based on keystroke logs, ranging from stud- ies on writing itself (Leijten and Van Waes, 2013; Bal- lier et al., 2018) to translation studies (Serbina et al., 2017; Carl et al., 2011) and language learning stud- ies (Chukharev-Hudilainen et al., 2019; Miller et al., 2008). Keystroke logs typically record rich behavioural infor- mation, such as pause duration between writing events and the speed of text sequence production, and Input- log also provides some levels of linguistic annotation for English and Dutch (tokenization, lemmatization, POS-tagging, chunking and syllabification ; see (Lei- jten et al., 2015) for more details). For most other languages (including French), full linguistic annota- tion of keystroke logs needs to be done as a sepa- rate step. Given the often non-canonical nature of the data, which contains phenomena similar to disfluencies and error correction encountered in spoken language (Gilquin et al., 2011), automatic annotation poses sim- ilar challenges to those encountered when processing transcribed spoken corpora, e.g. (Gerdes and Kahane, 2009). It is therefore not surprising that fully anno- tated keystroke log corpora remain rare (see (Serbina et …

Content truncated to 500 words