Show simple item record

dc.contributor.advisorBreslin, John
dc.contributor.authorVickey, Theodore
dc.date.accessioned2018-08-23T08:24:25Z
dc.date.available2018-08-23T08:24:25Z
dc.date.issued2018-06-13
dc.identifier.urihttp://hdl.handle.net/10379/7649
dc.description.abstractIn 2013, the World Health Organization coined the term “Globesity” to highlight the importance of the epidemic and impact as a major health problem in many parts of the world (World Health Organization, 2013). Many believe that technology has been a key to this decrease in physical and increase in sedentary behaviors. However, the effective use of technology may be one of the keys towards better health, fitness and wellness for the global population using a device they already use, a mobile phone. This research investigates a novel and scalable method of physical activity data collection through the use of Twitter. The publicly available fitness tweets can provide a wealth of demographic and activity information from Twitter users from around the world. Specifically, this dissertation reviews the use of five mobile fitness apps that allow participants to share their workouts from the app through Twitter. At the time of the data collection, this was the first known academic research in the world that combined Twitter, mobile fitness apps and physical activity. The underlying hypotheses and research questions explore if researchers can learn enough information to help decrease the incidence of physical inactivity from a population level. There were three research questions to be answered. The first was to determine if an automated data collector could be created and to then accurately identify fitness tweets shared from one’s mobile fitness app from the Twitter stream. If so, how can these fitness tweets be collected and processed and what are the limitations in the data processing of these fitness tweets? The second was to determine if an automated fitness tweet classification model could accurately quantify characteristics of physical activity such as but not limited to duration, type and intensity. If so, what is the process and what are the limitations in the collection of physical-activity minutes using Twitter? The third and final was to determine if additional demographic information from those who share their workouts online using Twitter be generated to give additional insights into the characteristics of the type of person whom fitness tweets? Building on existing research tools and literature on data collection, analysis and classification from computer science research, a Fitness Tweet Collection Tool and a Fitness Tweet Classification Model were created to process and catalogue over 2 million fitness tweets. The tools were designed in such a way that would allow future health and wellness researchers, without an in-depth knowledge of programming or coding, to modify the tools to conduct their own Twitter and health research. One such research project that used this data collection model was by researchers at Harvard studying skin cancer. Four hypotheses are presented from the resulting fitness tweets: that as one’s online influence increases, the busier they will be, thus less time to exercise; that women are more expressive in their tweets and would tend to fitness tweet more; that people feel better when they exercise, and this will be reflected in their tweets and that those who use mobile fitness apps are more interested in their physical activity, thus will report greater minutes of activity than those that were self-reported through surveys. The findings of this research and the three data experiments conducted from the collected data provide implications for not only future researchers, but also for mobile fitness developers. This research demonstrates that it is possible to collect self-reported health information, in this case physical activity, from a diverse population at different levels of physical activity from around the world.en_IE
dc.publisherNUI Galway
dc.subjectTwitteren_IE
dc.subjectMobile Fitness Appsen_IE
dc.subjectPhysical Activityen_IE
dc.subjectData Miningen_IE
dc.subjectEngineering and Informaticsen_IE
dc.subjectElectrical and Electronic Engineeringen_IE
dc.subjectData analyticsen_IE
dc.titleJust tweet it - The collection, processing, classification and analysis of 2 million fitness tweetsen_IE
dc.typeThesisen
dc.contributor.funderIrish Research Council for Science, Engineering and Technologyen_IE
dc.contributor.funderAmerican Council on Exerciseen_IE
dc.contributor.funderScience Foundation Irelanden_IE
dc.local.noteA collection of over 2 million physical activity related tweets over a six month period sent from one's smartphone using one of five different mobile fitness apps (Nike, RunKeeper, DailyMile, Endomondo and MyFitnessPal). From this dataset, a number of different experiments were conducted with the data to reflect self reported measures of physical activity and the sentiment of those activities in real time.en_IE
dc.local.finalYesen_IE
dcterms.projectSFI/08/CEI1380en_IE
nui.item.downloads6616


Files in this item

Attribution-NonCommercial-NoDerivs 3.0 Ireland
This item is available under the Attribution-NonCommercial-NoDerivs 3.0 Ireland. No item may be reproduced for commercial purposes. Please refer to the publisher's URL where this is made available, or to notes contained in the item itself. Other terms may apply.

The following license files are associated with this item:

Thumbnail

This item appears in the following Collection(s)

Show simple item record