Retrieving Compositional Documents Using Position-Sensitive Word Mover's Distance.
Martin TrappMarcin SkowronDietmar SchabusPublished in: ICTIR (2017)
Keyphrases
- spoken documents
- information retrieval
- word spotting
- word frequencies
- latent topics
- text queries
- text corpus
- document collections
- term frequency
- natural language text
- printed documents
- keywords
- relevant documents
- word pairs
- web documents
- related words
- multiword
- spoken document retrieval
- index terms
- concept space
- document space
- information retrieval systems
- document clustering
- user queries
- distance measure
- sentence level
- xml documents
- metadata
- document retrieval
- text documents
- retrieval systems
- indian languages
- test collection
- handwritten documents
- effective retrieval
- word similarity
- word frequency
- document representation
- linguistic information
- word co occurrence
- historical manuscripts
- related documents
- historical documents
- highly relevant documents
- search engine
- text classification
- stop words
- topic models
- document images
- query terms
- word recognition
- digital libraries
- relative position