Billions of Parameters Are Worth More Than In-domain Training Data: A case study in the Legal Case Entailment Task.
Guilherme Moraes RosaLuiz Henrique BonifacioVitor JeronymoHugo AbonizioRoberto de Alencar LotufoRodrigo NogueiraPublished in: CoRR (2022)
Keyphrases
- training data
- legal cases
- maximum likelihood
- domain specific
- decision trees
- domain independent
- parameter space
- training set
- classification accuracy
- parameter values
- parameter estimation
- logic programming
- active learning
- prior knowledge
- case study
- legal knowledge representation
- training instances
- legal reasoning
- data sets
- training dataset
- cross domain
- test bed
- training examples
- learning algorithm