Blending Topic-based Embeddings and Cosine Similarity for Open Data Discovery.
Maria Helena FranciscattoMarcos Didonet Del FabroLuis Carlos Erpen De BonaCelio TroisHegler TissotPublished in: ICEIS (1) (2022)
Keyphrases
- cosine similarity
- open data
- vector space
- distance measure
- linked data
- similarity measure
- open government data
- document clustering
- euclidean distance
- similarity function
- world wide
- vector space model
- low dimensional
- design choices
- tf idf
- knowledge discovery
- topic models
- similarity search
- semantic similarity
- co occurrence
- data analysis
- metadata
- semantic web