Project
Using Visual Analytics to Make Sense of Online Search Results
Despite the adoption of various resource discovery systems such as federated search engines, link resolvers, vertical search engines and the embracing of some Web 2.0 technologies, teaching staff and librarians find that students are not accessing high level academic materials and are mostly relying on the Internet to find information resources for their course work. The jointly commissioned JISC-BL CIBER report highlighted these issues as being common to the ‘Google Generation’.
Although technological solutions are a step in the right direction, they have some specific limitations: they do not allow for advanced searching or smart searching so true associations between keywords are not made. Additionally, bibliometric information is not drawn out nor made use of. Also, there is an obvious lack of a common standard in database platform design so not all solutions work in a systematic manner resulting in uneven and sometimes unreliable search results. The vertical search engines come up with a series of key words but often do not offer an intelligent analytical representation of the relationships between the words. What would be a very significant advance would be the development of a means by which these keywords could be displayed in ways that demonstrate the relative strength of their association – and which draw out, in a serendipitous way, additional unanticipated keyword links as proposed by the INVISQUE project.
The aim of the interactive visualization interface proposed by the INVISQUE project is to support sense-making, query formulation, and information search in resource discovery systems, by showing in visual representations, associations and relationships between the journal articles and/or books. This will support information search and retrieval tasks in and among large, loosely-coupled data sets, such as the library catalogue and disparate digital libraries and journal repositories. In addition, by enabling changes in viewing perspectives (e.g. rotation, re-ordering, re-collating) it will facilitate the chance discovery of unanticipated associations and resources. Other features could include readers to write reviews or questions.