Liquid Query: Multi-domain Exploratory Search on the Web


Published on

The presentation about Liquid Queries performed at WWW2010

Published in: Technology
  • Thanks for sharing this, I was always confused whenever I tried to do so, no matter how much I read or was consulted. This video was easy to understand and was thus VERY helpful.
    Are you sure you want to  Yes  No
    Your message goes here
  • Have a look at the demonstration video!

    <br /><object type="application/x-shockwave-flash" data=";hl=en_US" width="350" height="288"><param name="movie" value=";hl=en_US"></param><embed src=";hl=en_US" width="350" height="288" type="application/x-shockwave-flash"></embed></object>
    Are you sure you want to  Yes  No
    Your message goes here
No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • complex search is characterized by: multiple searches, possibly over multiple sessions and spanning multiple sources of information; a combination of exploration and more directed information finding activities; the need of note-taking, the variation of the search goal during the search process.
  • Liquid Query: Multi-domain Exploratory Search on the Web

    1. 1. Liquid QueriesMulti-domain Exploratory Search on the Web<br />Alessandro Bozzon, Marco Brambilla, Piero Fraternali, Stefano Ceri<br />WWW 2010, Raleigh (NC)<br />April, 29th 2010<br />
    2. 2. Context<br />The Web is a huge, heterogeneousdata source:<br />Unstructured information<br />Structured and semi-structured data<br />Linked Data<br />Documents are integrated with semi-structured or structured data and enriched with semantics<br />Search as a service<br />Google, Yahoo, Amazon, Twitter, YouTube, MySpace, and Facebook<br />Viability of systems based upon search service orchestration<br />
    3. 3. Problem Statement<br />Web queries get increasingly complexand specialized as the searcher’s expertise grows<br />Exploratory search<br />Multi-step search processes<br />From document searchto object/entity search<br />Multi-domain queries<br />multiple semantic fields of interest (e.g., travels, music, shows, food, movies, health, genetic diseases, etc.)<br />domain-specific (aka, vertical) search engines provide more focused results than general-purpose ones<br />the challenge is to match the results coming fromdifferentverticals<br />
    4. 4. An Example of multi-domain query<br />“… search for upcoming concertsclose to an attractive location (like a beach, lake, mountain, natural park, and so on), considering also availability of good, close-byhotels …” <br />Current approach the user can adopt:<br /><ul><li>Independently explore search services
    5. 5. Manually combine findings</li></li></ul><li>An Example of multi-domain query<br />“… expand the search to get information about available restaurantsnear the candidate concert locations, news associated to the event and possible options to combine further events scheduled in the same days and located in a close-by place with respect to the first one…”<br />
    6. 6. Related work<br />Exploratory Search: user’s intent is primarily to learn more on a topic of interest<br />Marchionini, G. Exploratory search: from finding to understanding. Communications ACM 49(4): 41-46 (2006)<br />“… exploratory search, that blends querying and browsing strategies, from retrieval that is best served by analytical strategies…”<br />Definition and analysis of the problem<br />White, R. W., and Drucker, S. M. Investigating behavioral variability in web search. 16th WWW Conf. (Banff, Canada, 2007)<br />Complex Search Vs. Exploratory Search<br />Aula, A., and Russell, D.M. Complex and Exploratory Web Search. ISSS: Information Seeking Support Systems Workshop (Chapel Hill, NC, USA, June 2008)<br />
    7. 7. Related work<br />Topic based search: instance of exploratory search centered on the goal of collecting information on a subject matter of interest from multiple sources<br /><ul><li>Kosmix: topic discovery engine, keyword search, a topic page summarizes the most relevant information on the subject
    8. 8. Hakia: resume pages for topics associated with user’s queries, natural language processing techniques</li></li></ul><li>Related work<br />Structured Object Search:process queries and present results that address entities or real world objects described in Web pages<br /><ul><li>Google Squared: keyword search, results collected in a table (called a square) featuring all the attributes relevant to the result items as columns headers
    9. 9. Google Fusion Tables: upload data tables (e.g., spreadsheet files) and join (or “fuse”) the data in some column with other tables</li></li></ul><li>The Liquid Query Approach<br />“ A new paradigm allowing users to formulate and get responses to multi-domain queries through an exploratory information seeking approach, based upon structured information sources exposed as software services…”<br /><ul><li>Compositeanswers obtained by aggregating search results from various domains
    10. 10. Highlightthe contribution of each search service
    11. 11. Joinof results based on the structural information afforded by the search service interfaces
    12. 12. Refinethe user query
    13. 13. Re-shape result list </li></ul><br />
    14. 14. Liquid Query Life Cycle<br />
    15. 15. Liquid Query Template<br /><ul><li>service interfaces: high level representation of a data provisioning service which comprises input parameters, output parameters, and ranking attributes
    16. 16. connection patterns for joining the involved service interfaces
    17. 17. additional selection or join predicates
    18. 18. a default ranking function defined over the scores of service interfaces
    19. 19. sorting and clustering options that can be applied on the extracted result set
    20. 20. the size for the result pages
    21. 21. a set of available query expansions towards further domains</li></ul>Example scenario: Initial query on Concert and Hotel objects; possible expansions on Events, related News, Restaurants; distance ranking for Concert, Restaurant and Hotel; date ranking on News and Photos <br />
    22. 22. Query Submission<br />Hotels<br />query conditions<br />Concert <br />query conditions<br />
    23. 23. Query Execution<br />
    24. 24. Result Exploration (SaaP - Search as a Process)<br /><ul><li>The current set of combinations is not satisfactory, so user asks for more values for a service (more one) or for all services (more all)
    25. 25. More concerts, more hotels, or more combinations
    26. 26. Add new information about further domains for selected combinations (expand)
    27. 27. Find close-by restaurants or co-located events
    28. 28. Aggregate information to ease analysis and readability (clustering, grouping)
    29. 29. Group events by venue
    30. 30. Reduce the number of shown items through filtering
    31. 31. Total walked distance for the night
    32. 32. Re-order (ranking or sorting)
    33. 33. Calculate derived values from existing ones
    34. 34. Total walked distance for the night
    35. 35. Alternative data visualization
    36. 36. Map, parallel coords, ...</li></li></ul><li> The back-end: YQL - Yahoo Query Language<br />A language and a platform that let Web application query, filter, and combine data from different sources across the Internet through SQL-like statements<br />Remote Query Interaction PrimitivesYQL commands + locally computed join operation<br />Local Resultset Interaction and Local Manipulation Primitives client side (GoogleGears DB) <br />Data Visualization Primitives mashups with visualization libraries<br />SELECT title, city, Rating.AverageRating, <br />latitude, longitude<br />FROM Local.Search (0,100)<br />WHERE (latitude,longitude) IN <br /> (SELECT latitude, longitude<br />FROM Upcoming.Events <br /> WHERE category_id=“1” AND<br /> min_date="2009-12-01" AND<br /> max_date="2009-12-31" AND<br /> location = "San Francisco" AND<br /> search_text=”Jazz<br />)AND query=“Hotel” AND radius = “0.5”<br /> AND Rating.AverageRating > 2<br />AND Rating.TotalRatings > 30<br />Flickr<br />Weather<br />Yahoo Upcoming<br />Yahoo Search<br />Amazon<br />Select / Create / Delete<br />NYTimes<br />Facebook<br />Music<br />Remote Filter, Sort, Limit<br />Local Filter, Sort, Limit<br />
    37. 37. Liquid Query Prototype Demo<br /><br />
    38. 38. Overview of Search Computing and Liquid Query<br />
    39. 39. Liquid Queries Prototype Architecture<br />Query, Table and Action Configuration<br />Local Data Configuration<br />LQ Widget Configuration<br />Query Configuration<br />YQL Service Configuration<br />
    40. 40. Liquid Query over Linked Data Demo<br />Example Scenario: Scientific Publication search<br /><br />
    41. 41. Conclusions and future work<br />Liquid Query<br /><ul><li>exploits the power of underlying vertical search services by joining and matching results
    42. 42. grants multi-domain exploratory search through a combination of search interfaces and data visualization
    43. 43. Details: “Search Computing – Challenges and Directions”Ceri, Brambilla (eds.), book by Springer, April 2010</li></ul>Future and ongoing works<br /><ul><li>thorough user evaluation studies
    44. 44. design of additional interaction and visualization options
    45. 45. implementation of visual tools for Liquid Query configuration
    46. 46. Integration of the Search Computing backend engine
    47. 47. Application of diversification and personalization to results</li></li></ul><li>Thanks! Questions?<br /><br /><br />