-
Notifications
You must be signed in to change notification settings - Fork 183
Description
Hi, I am interested in implementing an L2 Syntactic Complexity Analyzer for textstat. It automates the analysis of texts written by second language learners, which helps both researchers of L2 learners and L2 teachers in the classroom by providing 14 indices describing the syntactic complexity of the text.
The following calculations are necessary to implement the L2SCA:
- number of words
- ! number of clauses
- ! number of dependant clauses
- ! number of T-Units
- ! number of complex T-Units
- number of sentences
- ! number of coordinate phrases
- ! number of verb phrases
- ! number of complex nominals
In addition to the above, (Lu, 2010) uses the following tools to count them:
- ! A syntactic parser from which they derive syntax structure trees of each sentence in the text. Their choice was Stanford Parser
- ! Tregex to extract patterns from the trees
A ! denotes something that textstat does not currently offer, which would entail additional work in the process.
A web based version is hosted here: https://aihaiyang.com/software/l2sca/
Link to papers and the original, local implementation: http://www.personal.psu.edu/xxl13/downloads/l2sca.html
https://www.jbe-platform.com/docserver/fulltext/ijcl.15.4.02lu.pdf
Is this something we would like to add to textstat?