Empirical Study of LLM Fine-Tuning for Text Classification in Legal Document Review.
Fusheng WeiRobert KeelingNathaniel Huber-FlifletJianping ZhangAdam DabrowskiJingchao YangQiang MaoHan QinPublished in: BigData (2023)
Keyphrases
- empirical studies
- fine tuning
- text classification
- text documents
- document classification
- systematic review
- text classifiers
- term frequency
- automatic text classification
- text categorization
- topic discovery
- training documents
- viable alternative
- text mining
- empirical analysis
- fine tune
- bag of words
- real world data sets
- text data
- machine learning
- naive bayes
- document images
- document representation
- sentiment analysis
- document clustering
- multi label
- semantic features
- information retrieval
- feature selection
- retrieval systems
- sentiment classification
- classify documents
- data cleaning
- document collections
- uci datasets
- labeled data
- fine tuned
- legal knowledge
- logical structure
- tf idf
- knowledge discovery
- semantic information
- n gram
- information retrieval systems
- legal reasoning
- search engine
- electronic documents
- digital libraries
- topic models
- web documents