dc.contributor.author | Nizam, Naureen | |
dc.date.accessioned | 2015-04-08T14:08:03Z | |
dc.date.available | 2015-04-08T14:08:03Z | |
dc.date.issued | 2015-04-08 | |
dc.identifier.uri | http://hdl.handle.net/10222/56337 | |
dc.description.abstract | People often use navigation mechanisms such as menus, search, links and browser tools to find information within websites. The purpose of this research was to explore how user-generated content on social media can be used to help users find information within websites. In particular, we have examined the use of links (i.e., web page URLs) shared on social media sites, such as Twitter, and the use of this information to recommend relevant and popular web pages to the website visitors.
Our preliminary study explored how users use current navigation tools within websites. Our next study focused on the Twitter messages (“tweets”) to identify characteristics of websites that may benefit from the links shared on social media. Using Netlytic, we captured and analyzed tweets about four popular events. The results indicated that 25-47% of tweets across all four events contained a link to web pages. The majority of these web pages were several clicks deep (requiring more than two clicks from the home page). Based on these findings, we developed guidelines and a prototype - a Social Media Panel (SMP). This prototype displayed popular web pages as page thumbnails based on the aggregated information trending on social media sites.
A mixed methodological approach was followed for our final study which included a focus group and a user study. The focus group was used to solicit feedback on the prototype. The prototype was refined based on these findings and evaluated through a user study. The prototype was compared against the current navigation tools and we examined its effectiveness, efficiency and user engagement between the fact finding and browsing tasks.
Through questionnaires and semi-structured interviews, we concluded that participants found SMP to be effective, efficient and engaging for browsing tasks. The analysis of logs and participants’ on-screen activity, revealed that they performed the fact finding tasks faster than the browsing tasks. It was statistically proven that it took fewer clicks to complete the task using SMP. However, the use of SMP did not prove to make a significant difference in expediting the completion of the task.
The combined results from these studies provided a set of guidelines, and recommendations for the SMP. The research helped us develop a website link navigation model and refine the web information classification model for the two types of information seeking tasks: fact finding and browsing. We see the potential of this research to assist website visitors to discover and connect with other social media users who are interested in similar topics and eventually lead these users to topic driven online communities. | en_US |
dc.language.iso | en | en_US |
dc.subject | Website Navigation | en_US |
dc.subject | Social Media | en_US |
dc.subject | User Experience | en_US |
dc.title | Using Social Media Data to Improve Navigation within Websites | en_US |
dc.type | Thesis | en_US |
dc.date.defence | 2015-03-11 | |
dc.contributor.department | Faculty of Computer Science | en_US |
dc.contributor.degree | Doctor of Philosophy | en_US |
dc.contributor.external-examiner | Dr. Kelly Lyons | en_US |
dc.contributor.graduate-coordinator | Dr. Evangelos Milios | en_US |
dc.contributor.thesis-reader | Dr. Anatoliy Gruzd | en_US |
dc.contributor.thesis-reader | Dr. Bonnie MacKay | en_US |
dc.contributor.thesis-supervisor | Dr. Carolyn Watters | en_US |
dc.contributor.ethics-approval | Received | en_US |
dc.contributor.manuscripts | Not Applicable | en_US |
dc.contributor.copyright-release | Not Applicable | en_US |