Open-source Large Language Models are Strong Zero-shot Query Likelihood Models for Document Ranking.
Shengyao ZhuangBing LiuBevan KoopmanGuido ZucconPublished in: CoRR (2023)
Keyphrases
- document ranking
- language model
- relevance model
- query expansion
- probabilistic model
- language modelling
- document retrieval
- probabilistic retrieval models
- language modeling
- information retrieval
- relevant documents
- statistical language modeling
- n gram
- federated search
- expert search
- query terms
- query specific
- smoothing methods
- text retrieval
- pseudo relevance feedback
- original query
- relevance assessments
- retrieval model
- translation model
- information retrieval systems
- query processing
- passage retrieval
- vector space model
- relevance feedback
- retrieved documents
- retrieval effectiveness
- test collection
- ad hoc retrieval
- topic models
- web documents
- image retrieval