Twitter is among the most popular social media networking platforms. Properties of tweets (Twitter data) such as volume, content, context, velocity, spatial location of tweets, etc. are stochastic in nature. In this dissertation, we investigate a number of spatio-temporal statistical models for tweets.
The original contributions of this dissertation are as follows:
1. Fractal modeling of tweet volume revealed several dynamic hierarchical relationships. These mathematical relationships were derived for the hierarchical connections among the number of tweets, population of cities of origins, areas of these cities, etc. Studying the parameters of these models revealed several insights, such as: will a Twitter topic have a broader appeal across several geographically distributed cities, or will it have a limited influence concentrated over only a few cities.
2. Using a combination of spatial data clustering and partitioning, the geographic circles of influences of a Twitter topic and the source cities that influence the topic were identified. Data collected on a number of Twitter topics corroborates this observation.
3. A wavelet transform domain clustering for Twitter users' temporal activities is useful in modeling tweet rate forecasting. We propose a model that connects the tweeting rate and the urban population, which is useful in forecasting tweet rates. A data-driven comparison between the forecasting model and Holt-Winters method indicated that the proposed forecasting model has a lower forecasting error rate on average. Combining this forecasting model with the spatial model provided additional insights.
|Commitee:||Ganesan, Narayan, Man, Hong, Prasad, M.G., Subbalakshmi, K.P.|
|School:||Stevens Institute of Technology|
|School Location:||United States -- New Jersey|
|Source:||DAI-B 78/05(E), Dissertation Abstracts International|
|Keywords:||Social media, Spatio-temporal modeling|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be