BERTgrid: Contextualized Embedding for 2D Document Representation and Understanding.
Timo I. DenkChristian ReisswigPublished in: CoRR (2019)
Keyphrases
- document representation
- vector space
- bag of words
- vector space model
- document clustering
- document collections
- vector representation
- web documents
- data fusion
- document content
- text documents
- semantic information
- language model
- background knowledge
- data mining
- similarity search
- computer vision
- information retrieval
- feature vectors
- multiscale
- similarity measure