- Please refer to the Data Dictionary folder for detailed understanding of the features in all the datasets shared
- Development Data folder has both the datasets for Articles and Tweets. Please use this data to build your solution
- Sample Output folder has the structure of how the submission file should be like
- Sample Output 1 is for the Mobile_Tech Classification & Text Summarisation task while Output Sample 2 is for the Entity Based Sentiment Analsysis
- Non-Adherence to the structure of the sample outputs shared would lead to disqualification
- You can use the Headline Similarity Scores.ipnb notebook to calculate BERT, ROUGH and BLUE scores. This code will be used to evaluate the performance of the headline genertor.