A Novel Concept-based Search for the Web of Data using UMBEL and a Fuzzy Retrieval Model

Presented at: 9th Extended Semantic Web Conference (ESWC2012)

by Melike Sah, Vincent Wade

As the size of Linked Open Data (LOD) increases, the search and access to the relevant LOD resources becomes more challenging. To overcome search difficulties, we propose a novel concept-based search mechanism for the Web of Data (WoD) based on UMBEL concept hierarchy and fuzzy-based retrieval model. The proposed search mechanism groups LOD resources with the same concepts to form categories, which is called concept lenses, for more efficient access to the WoD. To achieve concept-based search, we use UMBEL concept hierarchy for representing context of LOD resources. A semantic indexing model is applied for efficient representation of UMBEL concept descriptions and a novel fuzzy-based retrieval model is introduced for categorization of LOD resources to UMBEL concepts. The proposed fuzzy-based model was evaluated on a particular benchmark (~10,000 mappings). These evaluation results show that we can achieve highly acceptable categorization accuracy and perform better than the vector space model.

Keywords: UMBEL concept hierarchy, categorization, concept-based search, data mining, fuzzy retrieval model, linked open data, semantic indexing

Resource URI on the dog food server: http://data.semanticweb.org/conference/eswc/2012/paper/research/174

Explore this resource elsewhere: