O'Neill, Frank (2016) Search Engine Results Comparison for Result Filtering. Masters thesis, Dublin, National College of Ireland.
PDF (Master of Science)
Download (2MB) | Preview
Search Engines are a highly complex mixture of technologies and business motives. Behind this complexity incentives to engineer the results for business motives or otherwise are sometimes made. One claim is that Google results are directed towards user search history or identity. Search results bubble towards user preferences. While the claim of another search engine DuckDuckGo is that user identity is not used to engineer search results. In this report two search engine results are compared DuckDuckGo and Google in a two way Google Signed In / Out configuration.
A new Google account is opened and a profile is built for this Signed In account with three specific search terms over a six week period. An automated web browser is then used to gather data for three configurations of - Google Signed In, Google Signed Out and Duckduckgo. The first thirty URL returns of each Search Term for each Search Engine configuration are then categorised using the Web Service Similarsites.com. The returned categories are then evaluated for filtering. Findings suggest Google Search has a difference between a Signed In and Signed Out Search. Divergence is stronger depending on the particular topic - a filter bubble indication.
The exercise has the limitation of reliance on similarsites.com integrity, a small time frame for analysis.
|Item Type:||Thesis (Masters)|
|Subjects:||Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
|Divisions:||School of Computing > Master of Science in Data Analytics|
|Depositing User:||CAOIMHE NI MHAICIN|
|Date Deposited:||30 Jan 2017 08:59|
|Last Modified:||30 Jan 2017 08:59|
Actions (login required)