We are provided a custom movie dataset in BIO format (Link: https://groups.csail.mit.edu/sls/downloads/movie/) which involves two database eng corpus with simple queries and trivia10k13 with complex. The task described in the given problem is generally defined as sequence tagging or sequence labelling in the literature.
The folder contains SL.ppt (Presentation file)
The code folder have the two ipynb file:
- Eng_Flair_Sequence_Tagging.ipynb : For Eng data
- Trivia10k13_Flair_Seqeunce_Tagging.ipynb : For Trivia10k13 data
The Model-Eng folder inside Code folder, contain log and test/dev ouput files of model trained on Eng data. The Model-Trivia folder inside Code folder, contain log and test/dev ouput files of model trained on Trivia10k13 data.
Data can be downloaded from : https://groups.csail.mit.edu/sls/downloads/movie/