Date of Award


Document Type


Degree Name

Master of Science (MS)

Legacy Department

Computer Engineering


Shen, Haiying (Helen)

Committee Member

Gemmill , Jill

Committee Member

Smith , Melissa C


Question and Answering (Q&A) systems are currently in use by a large
number of Internet users. Q&A systems play a vital role in our daily life as
an important platform for information and knowledge sharing. Hence, much
research has been devoted to improving the performance of Q&A systems,
with a focus on improving the quality of answers provided by users, reducing
the wait time for users who ask questions, using a knowledge base to provide
answers via text mining, and directing questions to appropriate users. Due
to the growing popularity of Q&A systems, the number of questions in the
system can become very large; thus, it is unlikely for an answer provider to
simply stumble upon a question that he/she can answer properly. The
primary objective of this research is to improve the quality of answers and to
decrease wait times by forwarding questions to users who exhibit an interest
or expertise in the area to which the question belongs. To that end, this
research studies how to leverage social networks to enhance the
performance of Q&A systems. We have proposed SocialQ&A, a social
network based Q&A system that identifies and notifies the users who are
most likely to answer a question. SocialQ&A incorporates three major
components: User Interest Analyzer, Question Categorizer, and Question-
User Mapper. The User Interest Analyzer associates each user with a vector
of interest categories. The Question Categorizer algorithm associates a vector of interest categories to each question. Then, based on user interest
and user social connectedness, the Question-User Mapper identifies a list of
potential answer providers for each question. We have also implemented a
real-world prototype for SocialQ&A and analyzed the data from
questions/answers obtained from the prototype. Results suggest that social
networks can be leveraged to improve the quality of answers and reduce the
wait time for answers. Thus, this research provides a promising direction to
improve the performance of Q&A systems.