TexShape: Information Theoretic Sentence Embedding for Language Models.
H. Kaan KaleHoma EsfahanizadehNoel EliasOguzhan BaserMuriel MédardSriram VishwanathPublished in: CoRR (2024)
Keyphrases
- information theoretic
- language model
- document level
- language modeling
- information theory
- mutual information
- probabilistic model
- theoretic framework
- retrieval model
- sentence retrieval
- n gram
- document retrieval
- information retrieval
- information theoretic measures
- information bottleneck
- language modelling
- query expansion
- mixture model
- test collection
- natural language
- context sensitive
- text summarization
- jensen shannon divergence
- query terms
- vector space
- pseudo relevance feedback
- jensen shannon
- statistical language models
- cross lingual
- document ranking
- multi modal
- language models for information retrieval
- machine learning
- bayesian networks
- smoothing methods
- text mining
- relevance model
- vector space model