Doctor of Philosophy, The Ohio State University, 2023, Computer Science and Engineering

The widespread adoption of Web 2.0 technologies and the proliferation of computing devices have allowed individuals to collaborate and interact with one another. However, it has also resulted in the collection of a massive amount of structured data in the form of graphs. Recently, network representation learning (NRL) methods have achieved tremendous success in machine learning tasks on graphs. However, there are still many challenges in this area that need to be addressed before the adoption of NRL methods in the real world. Specifically, these challenges include i. Identifying effective NRL methods from a huge pool of existing NRL methods for a specific machine learning task, ii. Scaling existing NRL methods on large graphs, and iii. Developing novel NRL methods that address the issues of heterogeneity and privacy. In our first contribution, we perform a systematic benchmarking of network representation learning (NRL) methods. Given the plethora of NRL methods, the challenge for practitioners is to identify which methods to choose for specific tasks. The lack of a standard assessment protocol and benchmark suite exacerbates this problem even further and often leaves practitioners wondering if a new idea represents a significant scientific advance. Our overall assessment – a result of a careful benchmarking of 15 methods for unsupervised network representation learning on 16 datasets (several with different characteristics) - is that many recently proposed improvements are somewhat of an illusion. Our benchmarking study revealed that most of the NRL methods cannot operate on large graphs. To address the scalability problem, we propose MILE, a multi-level framework that can scale existing NRL methods on large graphs in an agnostic manner with a little compromise on the representation learning quality. We next considered the problem of using representation learning for human mobility analysis. We find that existing NRL methods require human mobility data to (open full item for complete abstract)

Committee: Srinivasan Parthasarathy (Advisor); Shaileshh Bojja Venkatakrishnan (Committee Member); Huan Sun (Committee Member); Xiaoyi Raymond Gao (Committee Member) Subjects: Computer Engineering; Computer Science

Master of Science (MS), Wright State University, 2021, Computer Science

Link prediction in the domain of scientific collaborative networks refers to exploring and determining whether a connection between two entities in an academic network may emerge in the future. This study aims to analyze the relevance of academic collaborations and identify the factors that drive co-author relationships in a heterogeneous bibliographic network. Using topological, semantic, and graph representation learning techniques, we measure the authors' similarities w.r.t their structural and publication data to identify the reasons that promote co-authorships. Experimental results show that the proposed approach successfully infer the co-author links by identifying authors with similar research interests. Such a system can be used to recommend potential collaborations among the authors.

Committee: Tanvi Banerjee Ph.D. (Advisor); Krishnaprasad Thirunarayan Ph.D. (Committee Member); Michael L. Raymer Ph.D. (Committee Member) Subjects: Artificial Intelligence; Computer Science

Master of Science, The Ohio State University, 2019, Computer Science and Engineering

There is growing interest for learning representations for nodes in a network. Several embedding generation algorithms have been proposed in the last few years that generate high quality representations for downstream tasks like node classification and link prediction. NetMF is one such algorithm that provides the theoretical foundations for proving that several network representation learning techniques implicitly factorize a closed form matrix derived from the graph. However, the NetMF algorithm is slow and does not scale well, owing to the multiple dense matrix multiplication steps and Singular Value Decomposition (SVD). We present Fast-NetMF, a fast, highly scalable version of the NetMF algorithm with reduced running time. In this work, we investigate the acceleration of NetMF under single-GPU and multi-core CPU settings. We also investigate replacing the slow SVD based matrix factorization step for faster and more parallel-friendly factorization techniques like Non-negative Matrix Factorization (NMF).

Committee: Srinivasan Parthasarathy (Advisor); Sadayappan P (Committee Member) Subjects: Computer Engineering; Computer Science