Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

From multistage information seeking models to multistage search systems (IIiX 2014)

667 views

Published on

Presentation at Information Interaction in Context (IIiX) conference 2014. Best presentation award. Paper available via: humanities.uva.nl/~kamps/publications/2014/huur:from14.pdf

  • Be the first to comment

From multistage information seeking models to multistage search systems (IIiX 2014)

  1. 1. From Multistage Information - Seeking Models to Multistage Search Systems Hugo Huurdeman, Jaap Kamps! University of Amsterdam huurdeman@uva.nl, kamps@uva.nl ! ! ! Presentation at IIiX 2014, Regensburg
  2. 2. 1. Introduction • Search systems facilitate a rich diversity of tasks, from simple to complex ! • Information seeking models: • complex tasks involve multiple stages! • Search systems: • mainly one-size-fits-all approach Search macro inf. seeking stages micro search systems • Paper aim: • bridging gap between macro level information seeking models and concrete search systems
  3. 3. 2. Multistage Information ! Seeking Models macro perspective
  4. 4. 2.1 Information Seeking Models • Information seeking modeled in a multitude of ways: • as behavioral patterns (Ellis) • as nonlinear activities (Foster) • as problem-solving (Wilson) • as temporal stages (Kuhlthau), .. ! • Our main focus: • temporally based IS models • Kuhlthau [1991] (Vakkari [2001]) • cognitively complex (work) tasks • involving learning & construction information behavior information seeking information search [Wilson99]
  5. 5. 2.2 Kuhlthau: Information Search Process [1991] Initiation Selection Exploration Formulation Collection Presentation + uncertainty - feelings thoughts actions uncertainty optimism confusion clarity confidence (dis)satisfaction doubt direction vague focused seeking relevant information (exploring) seeking pertinent information (documenting)
  6. 6. 2.2 Vakkari’s adaptation (in [Vakkari01]) Prefocus Focus formulation Postfocus seeking general background information seeking specific information faceted backgr. information relevance hard to judge relevance easier to judge decrease of number of broader terms information sought relevance search terms increase of number of search terms, synonyms, narrower terms!
  7. 7. 2.3 Implications for design of search systems Formulation • Kuhlthau: • so far “considerable impact” LIS, “little impact” IR system design [Kuhlthau99] • “applicable concepts”: process, uncertainty, uniqueness, complexity, interest, .. Initiation Selection Exploration Collection Presentation Prefocus Focus Postfocus ! • Vakkari: • “more support needed in initial stages” [Vakkari00] • e.g. background information • potentially support evolving relevance, relevance criteria, search tactics, ..
  8. 8. 2.3 Implications for design of search systems • Observation: good general understanding of macro level inf. seeking stages, but hard to translate to concrete micro level system design choices. macro inf. seeking stages micro system design
  9. 9. 3. Search user interfaces ! supporting seeking micro perspective
  10. 10. 3.1 Search user interfaces supporting seeking • Search User Interface (SUI) design: • no straightforward task to design a UI with a high usability [Shneiderm05] ! • A (limited) number of available frameworks, guidelines and design pattern libraries for SUIs • e.g. M. Wilson’s framework of SUI features [Wilson11] Search Input Control Personalizable Informational
  11. 11. 3.2 SUI approaches: traditional search • Streamlined interfaces ! • Focus on • query formulation • result list inspection • Advantages: [Hearst09] • lower cognitive load • more accessible & understandable Highly optimized for lookup tasks, less for open-ended queries [Marchionini06]
  12. 12. 3.2 SUI approaches: exploratory search • Supporting open-ended inf. seeking ! • Support learning and investigation activities for complex information problems [Marchionini06] • Many potential exploratory SUI features [White09], e.g. • rapid query refinement, facets (input, control) • leveraging context, visualizations (informational) • histories/workspaces/task management (personalizable) [Yee03]
  13. 13. 3.2 SUI approaches: sensemaking & analytics • Support analysis & synthesis [Pirolli11] ! • Potential functions facilitating notetaking, hypothesis formulation & collaborative search [Hearst09] • some overlap with exploratory search [Hearst13]
  14. 14. [Dunne12]
  15. 15. 3.3 Implications for search stage support (1/2) • Different approaches in support: • streamlined interfaces (few features) • complex and analytical interfaces (many features) ! ! ! ! ! ! ! • Some relation • exploratory search and early stages Kuhlthau • sensemaking and intermediate/late stages Kuhlthau • Few interfaces explicitly support macro level search stages
  16. 16. 3.3 Implications for search stage support (2/2) • Observation: understanding of search system features at the micro level, but fragmented understanding of how they can support information seeking stages at the macro level macro micro inf. seeking stages search! system features
  17. 17. Reconciling macro and micro views • Idea: could we use the understanding of the information seeking models at the macro level to understand behavior at the micro level? macro micro inf. seeking stages search! system features
  18. 18. 4. Interface features and! search stage macro — micro perspective
  19. 19. 4.1 Interface features & search stage • Discussed before: Information seeking stages • effects on information sought, relevance, and search tactics. ! • Hypothesis: flow of interaction also influenced by stages Search
  20. 20. 4.2 Interface features & stage: previous work • Approaches: log analysis, eye tracking analysis ! • Use of specific SUI features in different stages • e.g. relevance feedback [White05], query suggestions [Niu14] • Use of various SUI features in different stages • e.g. [Kules09], [Kules12], [Diriye13] ! • Findings: • varying use(fulness) of interface features at different moments of (complex) search tasks • search stage sensitive and agnostic features
  21. 21. 4.3 An experiment • Experimental dataset [Tran & Fuhr, 2011] • ezDL Interface with rich feature-set • Amazon/LibraryThing book data • 12 participants • narrow, complex and self-defined search tasks • analysis: 3 complex tasks • mean task time: 11.4m ! • eyetracking and system data
  22. 22. 4.3 An experiment: findings 60%! 50%! 40%! 30%! 20%! 10%! 0%! 0,1! 0,2! 0,3! 0,4! 0,5! 0,6! 0,7! 0,8! 0,9! 1! Fixation Count %! Task Progression (0: start, 1: end)! DetailView! ResultView! QueryView! BasketView! ResultView#01-10! ResultView#11-20! 6! 5! 4! 3! 2! 1! 0! 0,1! 0,2! 0,3! 0,4! 0,5! 0,6! 0,7! 0,8! 0,9! 1,0! Count! Task Progression (0:start, 1:end)! BasketModi"cations! BasketMeanItems! • Complex task 1-3: Fixation count over time • Complex task 1: Basket use over time
  23. 23. 4.3 Findings: summary • Beginning: • significant focus on query (control feature) • low use basket (personalizable) • Middle: • increased detail inspections (informational) • increased basket activity • End: • changes tendencies result list (deeper items) & detail inspections • decreasing & increasing use basket near the end of the task 60%! 50%! 40%! 30%! 20%! 10%! 0%! 0,1! 0,2! 0,3! 0,4! 0,5! 0,6! 0,7! 0,8! 0,9! 1! Fixation Count %! Task Progression (0: start, 1: end)! DetailView! ResultView! QueryView! BasketView! ResultView#01-10! ResultView#11-20! 6! 5! 4! 3! 2! 1! 0! 0,1! 0,2! 0,3! 0,4! 0,5! 0,6! 0,7! 0,8! 0,9! 1,0! Count! Task Progression (0:start, 1:end)! BasketModi"cations! BasketMeanItems! 15! 10! 5! 0! 0,1! 0,2! 0,3! 0,4! 0,5! 0,6! 0,7! 0,8! 0,9! 1,0! Number of queries! Task progression (0:start, 1:end)!
  24. 24. 4.4 Implications for search stage support 1.No clear dichotomy between stages: gradual change feature use! • e.g. basic informational features, such as results list macro micro inf. seeking stages search! system features 2.However, some features useful at specific moments! • some input, personalizable features, such as query and basket mainly used in specific task stages • Some support for hypothesis: different flow of users’ atomic actions at different moments
  25. 25. 5. Conclusion and ! Discussion
  26. 26. 5. Conclusion and discussion • Information seeking behavior (macro level) & concrete design system (micro level) ! • Analysis suggesting different patterns of use in different stages macro micro inf. seeking stages search! system! design
  27. 27. 5. Conclusion and discussion • Open question: what is the best way to support different stages in search interfaces? Search Search 1. streamlined interface 2. complex interface
  28. 28. 5. Conclusion and discussion Search Search Search Stage 1 Stage 2 Stage 3 3. multistage interface • Third option: differentiating search stage sensitive features & offer customized SUI support for stages in complex search ! • by adaptively showing SUI features • by adjusting shown details of features • by changing prominence, position and size of features, …
  29. 29. Explorations: INEX Interactive Social Book Search [M. Hall, H. Huurdeman, M. Koolen, M. Skov, and D. Walsh (2014)]
  30. 30. 5. Future work • Integrating this approach into the user’s flow • without being intrusive or confusing ! • Ways to support evolving stages, e.g. • on the interface level (SUI feature selection & details) • on the system level (customized contents and ranking) ! • Guiding searchers in their complex search process Search
  31. 31. Acknowledgements • We are very grateful to Vu Tran and the Information Engineering Group at the University of Duisburg-Essen for allowing us to analyze data collected with the ezDL interface. ! • This research is supported by the Netherlands Organization for Scientific Research (NWO project # 640.005.001 - WebART)
  32. 32. References (1/2) • [Diriye13] A. Diriye, A. Blandford, A. Tombros, and P. Vakkari. The role of search interface features during information seeking. In TPDL, volume 8092 of LNCS, pages 235–240. Springer, 2013. • [Dunne12] C. Dunne, B. Shneiderman, R. Gove, J. Klavans, and B. Dorr. Rapid understanding of scientific paper collections: Integrating statistics, text analytics, and visualization. JASIST, 63 (12): 2351–2369, 2012. • [Hall14] M. Hall, H.C. Huurdeman, M. Koolen, M. Skov, and D. Walsh. Overview of the INEX 2014 interactive social book search track. CLEF 2014 Online Working Notes. Sheffield, United Kingdom.! • [Hearst09] M. A. Hearst. Search user interfaces. Cambridge University Press, 2009. • [Kuhlthau91] C. C. Kuhlthau. Inside the search process: Information seek- ing from the user’s perspective. JASIS, 42:361–371, 1991. • [Kuhlthau99] C. C. Kuhlthau. Accommodating the user’s information search process: challenges for information retrieval system designers. Bulletin of the ASIST, 25(3):12–16, 1999. • [Kules09] B. Kules, R. Capra, M. Banta, and T. Sierra. What do ex-ploratory searchers look at in a faceted search interface? In JCDL, pages 313–322. ACM, 2009. • [Kules12] B. Kules and R. Capra. Influence of training and stage of search on gaze behavior in a library catalog faceted search interface. JASIST, 63:114–138, 2012. • [Marchionini06] G. Marchionini. Exploratory search: from finding to understanding. CACM, 49(4): 41–46, 2006. …
  33. 33. References (2/2) • [Niu14] X. Niu and D. Kelly. The use of query suggestions during information search. IPM, 50:218– 234, 2014. • [Pirolli11] P. Pirolli and D. M. Russell. Introduction to this special is- sue on sensemaking. Human- Computer Interaction, 26:1–8, 2011. • [Shneiderm05] B. Shneiderman and C. Pleasant. Designing the user in- terface: strategies for effective human-computer interaction. Pearson Education, 2005. • [Tran11] V. T. Tran and N. Fuhr. Quantitative analysis of search sessions enhanced by gaze tracking with dynamic areas of interest. In Theory and Practice of Digital Libraries, volume 7489 of LNCS, pages 468–473. Springer, 2012. • [Vakkari01] P. Vakkari. A theory of the task-based information retrieval process: a summary and generalisation of a longitudinal study. Journal of Documentation, 57:44–60, 2001. • [Wilson99] T. D. Wilson. Models in information behaviour research. Journal of Documentation, 55:249–270, 1999. • [Wilson11] M. L. Wilson. Interfaces for information retrieval. In I. Ruthven and D. Kelly, editors. Interactive Information Seeking, Behaviour and Retrieval. Facet, 2011. • [White05] R. W. White, I. Ruthven, and J. M. Jose. A study of factors affecting the utility of implicit relevance feedback. In SIGIR, pages 35–42. ACM, 2005. • [White09] R. W. White and R. A. Roth. Exploratory search: Beyond the query-response paradigm. Synthesis Lectures on Information Concepts, Retrieval, and Services, 1:1–98, 2009.
  34. 34. From Multistage Information - Seeking Models to Multistage Search Systems Hugo Huurdeman, Jaap Kamps! University of Amsterdam huurdeman@uva.nl, kamps@uva.nl ! ! ! Presentation at IIiX 2014, Regensburg

×