Mode Effects’ Challenge to Authorship Attribution

Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics, pages 1146–1155 April 19 - 23, 2021. ©2021 Association for Computational Linguistics 1146 Mode Effects’ Challenge to Authorship Attribution Haining Wang1, Allen Riddell1, and Patrick Juola2 1Indiana University Bloomington, Bloomington, Indiana 2Duquesne University, Pittsburgh, Pennsylvania {hw56|riddella}@indiana.edu juola@mathcs.duq.edu Abstract The success of authorship attribution relies on the presence of linguistic features specific to individual authors. There is, however, lim- ited research assessing to what extent autho- rial style remains constant when individuals switch from one writing modality to another. We measure the effect of writing mode on writing style in the context of authorship attri- bution research using a corpus of documents composed online (in a web browser) and doc- uments composed offline using a traditional word processor. The results confirm the ex- istence of a “mode effect” on authorial style. Online writing differs systematically from of- fline writing in terms of sentence length, word use, readability, and certain part-of-speech ra- tios. These findings have implications for re- search design and feature engineering in au- thorship attribution studies. 1 Introduction That authorship attribution techniques work as re- liably as they do has been attributed to the fact that each individual has a distinctive writing style. Texts written by the same author can be recog- nized by analyzing lexical and syntactic features in documents (Juola, 2006). This principle is practi- cally successful in a variety of settings (Abbasi and Chen, 2008; Overdorf and Greenstadt, 2016; Afroz et al., 2014). In some cases, however, authorial style is challenging to detect. For example, autho- rial style fades as time goes by (Glover and Hirst, 1996; Baayen et al., 2002), varies considerably in collaborative environments (Graham et al., 2005; Kestemont et al., 2018; Zangerle et al., 2019), and drifts depending on document genre (Stamatatos, 2018; Koppel et al., 2007; Sapkota et al., 2016). The aforementioned changes are conspicuous due to the fact that there are certain markers indi- cating that a document may have been written in a fashion that will lead to stylistic variation. For instance, the presence of genre-specific words indi- cates a document may present a puzzle for standard authorship attribution techniques. Sometimes, how- ever, documents which may challenge an analysis of authorial style can be unannounced. This paper shows that authorial style changes with respect to sentence length, word use, readability, and certain part-of-speech ratios when the writing environment switches from traditional word processing software to an input box of a web browser. 2 Mode Effects Originally developed in survey research and ed- ucational testing, a “mode effect” describes the following phenomenon: a respondent may answer the same question differently depending on how a survey is administered (e.g., online vs. phone) (Hochstim, 1967; Leeson, 2006). Although dis- cussion continues about mode effects’ underlying mechanism (Kreuter et al., 2008; Sidi et al., 2017), contextualized magnitude (Carpenter and Alloway, 2019; Washburn et al., 2017), and adjustment meth- ods (Kolenikov and Kennedy, 2014), a consensus has been reached that …

Content truncated to 500 words