Lbl2Vec: An Embedding-based Approach for Unsupervised Document Retrieval on Predefined Topics.
Tim SchopfDaniel BraunFlorian MatthesPublished in: WEBIST (2021)
Keyphrases
- document retrieval
- information retrieval
- topic modeling
- document collections
- language model
- retrieval model
- relevance feedback
- ranking functions
- text retrieval
- cross language
- learning to rank
- topic models
- relevant documents
- passage retrieval
- xml retrieval
- document indexing
- document ranking
- spreading activation
- retrieved documents
- latent dirichlet allocation
- inverted index
- pseudo relevance feedback
- metadata
- information extraction
- machine learning
- supervised learning
- question answering systems
- query terms