Incorporating Text in Enterprise Information Systems

Enterprise Information Systems (EIS), the core ICT backbone of organizations, are based on structured data, which are stored in relational databases. These databases may contain text fields as attributes of objects, but lack functionalities to analyze text data. As a result, the considerable amount of valuable texts that is contained in enterprise systems' database cannot be exploited to enrich corporate activities and processes. At the same time, Natural Language Processing (NLP) techniques have been developed to analyze texts from other sources, such as emails and social media. However, these techniques fail to leverage upon the high quality additional information that is inherent in the structure or schema of the EIS database in order to improve their performance. In this paper, we reconcile the seemingly dichotomous worlds of EIS and NLP. We posit that our approach allows to enrich and incorporate text in enterprise systems.