Visualization of Ontology-Based Web Search

Type: 
PR/DA
Persons: 
1

Description

In web search, users retrieve satisfactory information by applying the most suitable set of search terms. This is a challenging task, especially for users new to a domain that have to undergo a cumbersome trial and error process to getting closer to their desired content. In this process small variations in search terms can have large impact on the set of query results.

The goal of this thesis is support the user in improving the relevance of her search query. This is done by visualizing how small changes to web queries affects the search results. Existing web ontologies from related domains are used to create variations of the user query to identify commonalities and differences in the results of these parallel searches. The text content of the search results then can be parsed to extract the most expressive key terms or phrases. An interactive visualization of the search history and its resulting key terms and phrases can make previous decisions transparent and suggest refined query terms.

This work will be done in collaboration with Institute for Innovation Management at the Faculty of Social Sciences, Economics and Business at the Johannes Kepler University in Linz.

Tasks

The main objective of the work is to design and implement an interactive, chronological visual representation of the sets of key terms summarizing the query results, how the terms overlap across multiple queries, and which terms are unique for a specific query variation.

Requirements

  • Strong interest in information visualization and human-computer interaction

  • Solid programming skills

  • Experience with web technologies (JavaScript, d3, ...) advantageous

  • Commitment to collaborate with experts from a different domain

  • Knowledge of English language (source code comments and final report should be in English)

Environment

The work should track and support search with an existing search engine within a standard web browser. The suggested environment is a Chrome extension. However, based on prior experiences of the student, alternative environments can also be considered. 

Contact

For more information please contact Manuela Waldner (waldner@cg.tuwien.ac.at).