A Survey on Knowledge Extraction Techniques for Web Tables

Web tables are worthy sources of relational information. The number of high-quality tables with useful relational information is rapidly increasing to hundreds of millions. Some search engines usually ignore meanings of entities and relationships in indexing thus they have poor performance in tabular data to a suitable field of research is the transformation of web tables into machine-readable knowledge. We first study overview of the use of web tables in different domains then focus on understanding knowledge of web tables. The results indicate that by combining old Information Extraction techniques, and table features and general inference models can extract Knowledge from web tables.