Results announced
The task has two parts
Task A : Binary Classification: Identify whether a given sentence contains a causal event (either cause/effect) :
Task B: Tagging : Annotate each word in a sentence in terms of the four labels cause (C),effect(E), causal connectives(CC) and None:
Such as shown in the figure below
The tagset is : cause (C), effect(E), Causal Connectives (CC) -, if a word is unmarked it will be assumed as None
1. Heavy rains during the harvest may lead to the rotting of onion production.
Annotations: Heavy\Cause rains\Cause during the harvest may\causal connective lead\causal connective to\causal connective the rotting\Effect of\Effect onion\Effect production\Effect.
2. From southern states the new crop of onions got ruined due to heavy rainfall.
Annotations: from southern states the\Effect new\Effect crop\Effect of\Effect onions\Effect got\Effect ruined\Effect due\connective to\causal connective heavy\Cause rainfall\Cause.
3. Most of the people fear that the NRC implementation will cause a rush of immigrants that may alter their demographic and linguistic uniqueness.
Annotation: Most of the people fear that the NRC\Cause implementation\Cause will\causal connective cause\causal connective a rush\Effect of\Effect immigrants\Effect that\Effect may\Effect alter\Effect their\Effect demographic\Effect and\Effect linguistic\Effect uniqueness\Effect.
In practical scenarios, there can be nested cause and effect embedding like the second example sentence below. However, for out training and test set, we have kept only one cause and effect relation per example instance.
We have a test dataset annotated which will not be released during the training phase. We will post this held out dataset without target labels during the evaluation phase and ask the participants to submit their results on this test dataset. To prevent manual interference in final results, the participants will be asked to share the code for running their models on the final test dataset along with results.
Note: Results of Task A and Task B will be separately evaluated, therefore, participants need to submit the runs and results separately.
Task B will be annotated at the word level as the tagging is done at the word level.