Efficient High-Level Semantic Enrichment of Undocumented Enterprise Data

Abstract

In absence of a data management strategy, undocumented enterprise data piles up and becomes increasingly difficult for companies to use to its full potential. As a solution, we propose the enrichment of such data with meaning, or more precisely, the interlinking of data content with high-level semantic concepts. In contrast to low-level data lifting and mid-level information extraction, we would like to reach a high level of knowledge conceptualization. Currently, this can only be achieved if human experts are integrated into the enrichment process. Since human expertise is costly and limited, our methodology is designed to be as efficient as possible. That includes quantifying enrichment levels as well as assessing efficiency of gathering and exploiting user feedback. This paper proposes research on how semantic enrichment of undocumented enterprise data with humans in the loop can be conducted. We already got promising preliminary results from several projects in which we enriched various enterprise data.