jjzha / thesis-script

Code base of my bachelor thesis named "Labeled Bilexical Dependencies for Deception Detection".

Provided are 4 directories:

• csicorpus: the CSI corpus of Verhoeven and Daelemans (2014), in the vanilla state.
• features: every feature has a directory with 2 files (features.txt, labels.txt), preprocessed and such.
• parsed_reviews: the raw reviews parsed by the Alpino Parser (all, positive, and negative data).
• tokenized_reviews: tokenized sentences for the Alpino Parser using sentokenizer.py

Provided are 3 scripts:

• classification.py: main script used for classification tasks.
• sentokenizer.py: script to tokenize the sentences, appropriate for the Alpino Parser.
• parser.py: script used to parse the tokenized sentences.

If anything is unclear please contact me: [e-mail]