Observlets: Empowering Analytical Observations on Web Observatory

Aastha Madaan, Thanassis Tiropanis, Srinath Srinivasa, Wendy Hall

Abstract: The Web observatory is proposed as a global catalogue for sharing data-sets and analytic applications to support researchers from a variety of disciplines for analysing huge amount of research data for Web Science research. However, often these users fail to understand various transformations and consequences of complex data processing involved in a data analytic application. Therefore, there is a need to enable these users develop and re-use analytic applications on web observatory. In this study, we propose formal design patterns called "Observlets" for analytic applications to "observe" various web phenomena. The observlets provide abstract definitions for intermediate analysis required for a data analytic application. The users can share observlets across distributed web observatory nodes. The observlets are aimed to enhance end-users' awareness and engagement on web observatory and support programmers for innovating various data analytic applications.