Text analytics and document discovery in heterogeneous data sources
Nowadays, medium and large enterprises consume and generate information from several data sources and services. When choosing the tool for processing and storing this information, decisions are typically taken based on the needs of the involved teams. Although this strategy increases the efficiency of individual teams, the collaboration between team and departments may be hindered by such heterogeneity. As an example, a typical customer support team must daily consume data from: their own case tracker, for assessing if the current customer problem is already solved; wikis, Knowledge Bases and PDFs, for consulting technical documentation; and online forums, as customer and engineers may informally propose solutions on those channels. Besides the technical peculiarities of each system, the language and level of details are different due to the distinct objectives of the actors involved in the generation of such data. Such humongous amount of information reduces the efficiency of customer support teams and, hence, the customer satisfaction.
EULER aims at tackling such challenge by studying text analytics techniques for extracting information from documents and enabling the discoverability of links between documents across the different and heterogeneous data sources.