LINKED OPEN DATA AND RESEARCH KNOWLEDGE MANAGEMENT: AN EXPLORATORY SEARCH AND VISUALIZATION FRAMEWORK

In a relatively short time span people have moved from an analog society of paper and telephone to a society that is converting the entire world into a digital landscape. As a result, the amount of data available online today is overwhelming. The amount of data, its sheer size, often inhibits researchers, policymakers or ordinary people to obtain insight in these data in an efficient and effective way. Hence the need for solutions that enable users to discover, explore and analyze data. This paper describes and evaluates an exploratory search and visualization framework that enables non-expert users to interact and intuitively explore academic and bibliographic information. The framework, called the ‘Linked Open Data Vizualizations Suite’ (LOD/VizSuite), uses the Linked Open Data (LOD) of RILOD, a database with academic bibliographic data from heterogeneous sources on research in Flanders (northern part of Belgium), and supports exploratory search and data analysis by means of interactive graph-based visualizations.

In the first part of the paper the stage is set for describing LOD/VizSuite by theoretically unpacking concepts such as ‘open data’ (data that can be freely used, reused and redistributed by anyone – subject only, to the requirement to attribute), ‘linked data’ (structured data so that it can be interlinked), and ‘big data’ (a collection of data sets so large and complex that it becomes difficult to process using on-hand database management tools). We also elaborate on the differences between exploratory search and exploratory data analysis, an approach that allows the data itself to reveal its underlying model and relationships.

After reviewing the literature and describing the state of the art in the field of exploratory search and LOD, the second part of the paper describes LOD/VizSuite, its goals and its main user affordances. Screenshots of the user interface as well as a hyperlink to a demonstrator application are provided.

In the third part of the paper we determine the impact and quality of the visualization tool and analyze how the test users (n=40) explored and perceived the visualizations, using a multi-method approach including observations, experiments and questionnaires. Preliminary results indicate that a flow of graph-based visualizations is feasible with Linked Open Data and that users perceived that they found relevant insights about the researchers or disciplines they were exploring or looking for. Test users declared that, although not used to explore information in this way, they found the visualizations interesting. Still, others stated that they were sometimes confused, mainly due to the incremental appearance of the visualized information. We conclude that search and visualization frameworks on Linked Open Data and LOD/VizSuite in specific, can be supportive for scholars in academia as well as for policymakers in the research field.

This paper contributes to the field of knowledge management, open data and data visualization by empirically assessing the effectiveness and efficiency of LOD visualization technology, by theoretically unpacking and discussing key concepts, and by describing a case study analysis of LOD/VizSuite.