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.

Model driven crowdsourcing of search (CrowdSearch2012 workshop at www2012)


Published on

Even though search systems are very efficient in retrieving world-wide information, they can not capture some peculiar aspects and features of user needs, such as subjective opinions and recommendations, or information that require local or domain specific expertise. In this kind of scenario, the human opinion provided by an expert or knowledgeable user can be more useful than any factual information retrieved by a search engine.
In this paper we propose a model-driven approach for the specification of crowd-search tasks, i.e. activities where real people – in real time – take part to the generalized search process that involve search engines. In particular we define two models: the“Query Task Model”, representing the meta- model of the query that is submitted to the crowd and the associated answers; and the “User Interaction Model”, which shows how the user can interact with the query model to fulfill her needs. Our solution allows for a top-down design approach, from the crowd-search task design, down to the crowd answering system design. Our approach also grants automatic code generation thus leading to quick prototyping of search applications based on human responses collected over social networking or crowdsourcing platforms.

Published in: Technology
  • Login to see the comments

Model driven crowdsourcing of search (CrowdSearch2012 workshop at www2012)

  1. 1. A Model-Driven Approach for Crowdsourcing Search Alessandro Bozzon, Marco Brambilla, Andrea Mauri Politecnico di Milano Contact marcobrambiCrowdSearch 2012 workshop @ World Wide Web Conference (WWW2012), Lyon, April 17th, 2012
  2. 2. Outline• Rationale• (Meta)Models• Application• Demo• Outlook
  3. 3. SW Models + Social + Search =MD CrowdSearch
  4. 4. Rationale: increasing quality in exploratory search• From exploratory search to friends and experts feedback• Emphasis on social relations more than anonymous crowds Initial query Exploration Exploratory step Human Search Search System System Exploration step System API Social API Database / Crowd / IR index Community
  5. 5. Example
  6. 6. Deployment: Advantages of MDD• Multiple social platform deployment Generated query template Embedded External Standalone application application application API Social/ Crowd platform Native Embedding behaviours Community / Crowd
  7. 7. Search task management problemsTask splitting: the collection is too complex relative to the cognitive capabilities of users.Task structuring: the task is too complex or too critical to be executed in one shot.Task routing: a task can be distributed according to the values of some attribute of the collection.User interaction: search tasks may imply complex UI design• Easy to address through a model-driven approach
  8. 8. Efficient development of CrowdSearch appsApply model-driven techniques to Social and Search: MacroTask Description (BPMN) M2M Transformation MicroTask Description (BPMN) M2M Transformation User Interaction Model (WebML+ER) M2T Transformations Stand-alone Application embedded application in social network
  9. 9. Model extensions for Social BPMProcess and applications models are extended to (task- or incorporate social issues: login, post, tag, rate, share, ... Platform- specific) Social Process Model Social Application Model Comment Vote It is used to define: It is used to define: •Exchange of user profiles from/to SN •Social actors (e.g., Community Pools) •Social data (e.g., shared content) •Social Activities (twittering, voting, following..) •Interface and components for social tasks (e.g., •Social events twittering, voting, tagging, following) Based on BPMN social design patterns Based on WebML social components
  10. 10. The content (meta)model Field 1 N Schema CrowdObject N 1 Relation N 1 Outgoing From type: String name: String N 1 type: String 1 1 name: String N Incoming To idField N 1 1 FieldInstance User value: String Input Output 1 N N 1 user: String Answer password: String N 1 N email: String N 1 Query • Like N • Add question: String Responder Asker • Comment type: String 1 N • Modify open: boolean • …
  11. 11. WebML models – question definition UI model• user interaction + integration with social platformModel for defining a question:
  12. 12. WebML models – Response UI model
  13. 13. Rendering of the application (summary)
  14. 14. Model Driven Engineering of SocialSearch applications• WebRatio (, MDD tool that manages app developmentin three steps: Design Customize Generate the Model the Rules the Application• MDD Tools enable: fast prototyping, multi-platform deployment, model-driven debugging, and early assessment of alternative strategies
  15. 15. Social experiments and quantitative evaluations• See you on Friday, for the full paper presentation:Answering Search Queries with CrowdSearcher
Alessandro Bozzon, Marco Brambilla, Stefano Ceri
  16. 16. ReferencesThanks! • • • www.cubrikproject.euQuestions? • Contact: Marco Brambilla marcobrambi