Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Computer Engineering and Computer Science

Degree Program

Computer Science and Engineering, PhD

Committee Chair

Elmaghraby, Adel

Committee Co-Chair (if applicable)

Rouchka, Eric

Committee Member

Rouchka, Eric

Committee Member

Hieb, Jeffrey

Committee Member

Losavio, Michael

Committee Member

Imam, Ibrahim

Author's Keywords

natural language processing; transfer learning; lifelong learning; sentiment classification; machine learning; adaboost


The idea of developing machine learning systems or Artificial Intelligence agents that would learn from different tasks and be able to accumulate that knowledge with time so that it functions successfully on a new task that it has not seen before is an idea and a research area that is still being explored. In this work, we will lay out an algorithm that allows a machine learning system or an AI agent to learn from k different domains then uses some or no data from the new task for the system to perform strongly on that new task. In order to test our algorithm, we chose an AI task that falls under the Natural Language Processing domain and that is sentiment analysis. The idea was to combine sentiment classifiers trained on different source domains to test them on a new domain. The algorithm was tested on two benchmark datasets. The results recorded were compared against the results reported on these two datasets in 2017 and 2018. In order to combine these classifiers’ predictions, we had to assign these classifiers weights. The algorithm made use of the similarity between domains when inferring the weights for the classifiers trained on the source domains by measuring the similarity between these source domains and the domain of the new task concluding, that domain similarity could be used in computing weights for classifiers trained on previous tasks/domains.