Enabling Opinion-Driven Decision Making!

Opinions carry a wealth of information! You have opinions within blogs, e-commerce sites, user review sites, expert review sites, social networking sites, and in many other sources. These unstructured, scattered and disorganized opinions when seen individually, seem very ordinary. However, as we aggregate, summarize and leverage these opinions in creative ways, these opinions can become powerful decision making tools for both users and organizations. For users, this means decision making becomes quicker and more informed, making you an empowered user. For companies, this entails improved user retention, improved conversion rates and happier customers. FindiLike Hotel Search showcases a suite of research driven technologies to help transform opinions into powerful decision support tools.


FindiLike Hotel Search

FindiLike Hotel Search is a new generation search and review (opinion) analysis system that helps users find hotels based on their opinion preferences along with other common preferences. The opinion preferences can be expressed as a set of natural keywords such as 'close to place xyz', 'safe neighborhood', 'good breakfast', etc. In addition, the system also provides user review summarization tools to further analyze and narrow down on hotels in consideration. User productivity is significantly improved as you only have to select a hotel from a handful of choices.

Key Features

  • Search for hotels based on any type of unstructured preferences - pet friendly, clean, friendly staff, swimming pool, etc.
  • Search for hotels based on distance from a landmark or address
  • Rank hotels based on unstructured opinion preferences and structured preferences
  • Well-formed summarization of user reviews
  • Multi-word navigational word cloud of user reviews
  • Search within user reviews

    Tour

    FindiLike Product Demo from Kavita Ganesan


    Screenshots:

    [ See results on FindiLike Hotel Search system ]

    Hotels within 5 miles of University of Arizona - said to be 'pet friendly' and 'clean'

    Bigger image

    Multi-word buzz phrases generated from reviews of the Radisson Suites Tucson

    Bigger image

    Well-formed opinion summaries for the Radisson Suites Tucson

    Bigger image

    [ See results on FindiLike Hotel Search system ]

    Related Publications:

  • IR Journal - Opinion-based entity ranking...
  • WWW '12 - Micropinion generation: An Unsupervised Approach to...
  • WWW '12 - FindiLike: preference driven entity search...
  • COLING '10 - Opinosis: A Graph Based Approach To...

    Contact

    FindiLike Hotel Search was developed by Kavita Ganesan and ChengXiang Zhai, Professor of Computer Science at the University of Illinois at Urbana Champaign. This initiative started off as a research project to showcase the power of an Opinion-Driven Decision Support System. Some of the internal algorithms for Opinion Summarization are currently being used by other systems and researchers. The FindiLike system is currently being extended by students and is used for both evaluation of new algorithms as well as helping actual users to easily find hotels.

    If you have questions about using our technology (or parts of it) for your data or within your own application, please e-mail info@findilike.com to find out more. If you have technical questions please contact kganes@findilike.com