You are here:
Home › Final Year Project List ›
Understanding and Visualizing Search Networks

Understanding and Visualizing Search Networks

People
Project Mentor: 
Barry Smyth
Student: 
Kenneth O'Brien
Project Specification
Subject: 
Web Search
Project Type: 
Applied Research
Software Requirements: 
Java, Javascript, PHP, AJAX
Hardware Requirements: 
Mac or PC
Preassigned: 
Yes
Description
General Information: 

HeyStaks is a social search engine that adds 'sharing' to traditional search infrastructure. Using the HeyStaks browser toolbar, searchers can create and share so-called 'search staks', helping them to organise their searches in to meaningful topics, and share their search experiences with others. In this way HeyStaks leads to the creation of 'search networks', networks of individual searchers who actively share search experiences in much the same way that other social networks share their favourite photographs, or their activities, or their contact-lists.

In this project the student will look at various ways to help users make the most of their search networks and will actively explore different ways to visualize and harness search experiences. For example, a primary task will be to develop an efficient browser-based visualization tool that will allow a HeyStaks user to picture their extended search network and help them explore their search network in a highly interactive fashion.

This project is challenging and will suit students who have very strong programming skills, and who are confident that they can quickly come to grips with a range of programming technologies including Java, Javascript, AJAX, and PHP.  It will involve a mixture of Web programming and advanced computer graphics.

Mandatory: 
  • Develop an efficient representational format for manipulating search networks. This will require integration with the HeyStaks back-end database and the ability to extract real-time information from its search networks. This is a fairly straightforward information extraction/database integration task that will rely on a range of standard web programming technologies.
  • Evaluation of existing graph visualization techniques and libraries. This will involve significant background research and the ability to install and experiment with complex third-party software in various stages of release.
  • Develop a practical (i.e., efficient) graph visualization technique capable of rendering search networks of up to 100 nodes in real-time. This is challenging in and of itself. This requires some complex programming and the ability to craft efficient and manageable code. It is likely that the student will explore a force-directed visualization technique.

 

Discretionary: 
  • Develop an enhanced graph visualization technique capable of rendering search networks of up to 500 nodes in real-time.
  • Develop a variety of tools for interacting with these visualizations (pan, zoom, select etc.)
  • An evaluation of the evolution of search networks; this is considered to be a parallel activity that will allow the student to explore some of the more empirical questions relating to search networks and how they evolve within a social search system like HeyStaks.
Exceptional: 
  • Explore a variety of next-generation web-based animation techniques to develop a graphically rich visualization experiences.
Reading: 

A wide variaty of recent publications on collaborative/social web search are available online. Interested parties can be provided with sample papers upon request. To get started:

  • Smyth, B., Briggs, P., Coyle, M., and O’Mahony, M. P. A case-based perspective on social web
    search. In ICCBR (2009).
  • Smyth, B., Briggs, P., Coyle, M., and O’Mahony, M. P. Google? shared! a case-study in social
    search. In User Modeling, Adaptation and Personalization (June 2009), Springer-Verlag.
  • Church, K., Smyth, B., and Oliver, N. Visual interfaces for improved mobile search. In Visual
    Interfaces to the Social and the Semantic Web (VISSW 2009) (Florida, USA, January 2009).
  • Smyth, B. A community-based approach to personalizing web search. IEEE Computer 40, 8 (2007),
    42–50.

In addition interested students can find many tutorials and guides to cover the core technologies mentioned above. A repeated word of caution: students should not view this project as a good way to learn about Javascript/PHP/AJAX, unless they are supremely confident that they can pick these technologies up quickly; the main focus of the project is on using these technologies to develop visualization solutions.