To be filled out by faculty nominator, can be edited by student:
Dr. Elise DeDoncker
Most existing sentiment classification methods for social media focus on document-level classification. They utilize local text information and ignore the crucial characteristic information of users. These methods usually suffer from high model complexity and only exhibit word-level preference.
Therefore, motivated by the successful utilization of deep neural networks in computer vision, speech recognition and natural language processing and their ability of learning in multi-prospective ways, a neural network based sentiment analysis model is proposed to incorporate user-level information into sentiment classification.
By user-level information we mean information extracted or inferred from the user that helps the proposed model to engage in deep learning about users, which in turn improves its performance. The extraction of such information is a challenging task that will be explored in detail through this research.
WMU ScholarWorks Citation
Alharbi, Ahmed Sulaiman M, "Enhance a Deep Neural Network Model for Twitter Sentiment Analysis by Incorporating a User-Level Information" (2017). Research and Creative Activities Poster Day. 221.
Available for download on Wednesday, April 24, 2019