User Exploration of Search Space Using Tradeoffs
We describe a system for representing search results as an interactive graph, in contrast to the common static representation as an ordered list. The search terms are represented as Key-Nodes, and the results are represented as connected Record-Nodes. The user can then explore tradeoffs by assigning different importance values to Key-Nodes. We suggest algorithms for placement of Record-Nodes in the graph to ensure both smooth change of the representation in response to changes in user preferences, and clustering of results. The work was implemented using Haskell and a client-side Web interface.
KeywordsSearch Graph Graph representation Tradeoff
- Page, L., Brin, S., Rajeev, M., Terry, W.: The PageRank Citation Ranking: Bringing Order to the Web. Stanford University, Technical report (1998)Google Scholar
- Hearst, M.A.: Search User Interfaces. Cambridge University Press, (2009). Also available on the web http://searchuserinterfaces.com/book/
- Stolper, C.D., Kahng, M., Lin, Z., Foerster, F., Goel, A., Horng, D.: GLOs: Graph-level operations for exploratory network visualization. In: CHI 2014 Extended Abstracts on Human Factors in Computing Systems, pp. 1375–1380. ACM (2014)Google Scholar
- Dwyer, T., Mears, C., Morgan, K., Niven, T., Marriott, K., Wallace, M.: Improved optimal and approximate power graph compression for clearer visualization of dense graphs. In: Pacific Visualization Symposium (PacificVis), pp. 105–112. IEEE (2014)Google Scholar