learned_hands_consumer

Classify if a user post implicates legal isssues related to consumer.

---

Source: Learned Hands

License: CC BY-NC-SA 4.0

Size (samples): 620

Legal reasoning type: Issue-spotting

Task type: Binary classification

Task description

This is a binary classification task in which the model must determine if a user’s post discusses issues people face regarding money, insurance, consumer goods and contracts, taxes, and small claims about quality of service.

Task construction

This task was constructed from the LearnedHands dataset. Please see their website for more information on annotation. Our task consists of a binarized version of the original dataset, with “negatives” randomly sampled from posts with other topics. This dataset is class balanced.