Visual Trend Analytics in Digital Libraries

The early awareness of upcoming trends in technology enables a more goal-directed and efficient way for deciding future strategic directions in enterprises and research. Possible sources for this valuable information are ubiquitously and freely available in the Web, e.g. news services, companies’ reports, social media platforms and blog infrastructures. To support users in handling these information sources and to keep track of the newest developments, current information systems make intensively use of information retrieval methods that extract relevant information out of the mass amount of data. The related information systems are commonly focused on providing users with easy access to information of their interest and deal with the access to information items and resources [1], but they neither provide an overview of the content nor enable the exploration of emerging or decreasing trends for inferring possible future innovations. The gathering and analysis of this continuously increasing knowledge pool is a very tedious and time-consuming task and borders on the limits of manual feasibility. The interactive overview on data, the continuous changes in data, and the ability to explore data and gain insights are sufficiently supported by Visual Analytics and information visualization approaches, whereas the appliance of such approach in combination with trend analysis are rarely propagated. In fact, these so-called early signals require not only an analysis through machine learning techniques to identify emerging trends, but also human interaction and intervention to adapt the parameters used to their own needs [2]. There are two main aspects to consider in the analysis process: 1) which data reveal very early trends and 2) how can human be involved in the analysis process [3].