Expression Establishment

This release contains a new measure, expression establishment, proposed by Yuya Asano, Diane Litman, Mingzhi Yu, Nikki Lobczowski, Timothy Nokes-Malach, Adriana Kovashka, and Erin Walker. 2022. Comparison of Lexical Alignment with a Teachable Robot in Human-Robot and Human-Human-Robot Interactions. In Proceedings of the 23rd Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 615–622, Edinburgh, UK. Association for Computational Linguistics.

Alpha release of considering only specified lexicons

v1.1-alpha

This is meant to consider only shared and self-repeated lexicons that contain certain words researchers may be interested in (e.g. task-related words). For example, if you specify "Alice" and "book", only shared and self-repeated lexicons that contain either "Alice" and "book" in them are considered when computing the alignment measures.

Updates

Now you can specify the lexicons you want to consider with the "-l" option for both offline and online versions.
For example, if you have a file alice-vocab.txt (it does not have to have .txt extension),

java -jar dialign.jar -i input-directory/ -o output-directory/  -l alice-vocab.txt

for offline usage, and

java -jar dialign-online.jar -f excerpt-alice-partial.tsv -o dialign-online-excerpt-alice-partial.tsv -l alice-vocab.txt

for online usage.

Formatting the lexicon file

It must contain one lexicon (could consist of multiple words) per line with no following whitespace except for the newline.