04/09/2021
By Karen Volis

The Kennedy College of Sciences, Department of Computer Science, invites you to attend a doctoral dissertation defense by Yuting Xie on "Financial Knowledge Graph Construction, Querying, and Indexing."

Ph.D. Candidate: Yuting Xie
Defense Date: Tuesday, April 20, 2021
Time: 1 p.m. EST
Location: This will be a virtual defense via Zoom. Those interested in attending should contact yuting_xie@student.uml.edu at least 24 hours prior to the defense to request access to the meeting.

Committee Chair (Advisor): Tingjian Ge, Professor, Computer Science Department, University of Massachusetts Lowell
Committee Members:

  • Cindy Chen, Associate Professor, Computer Science Department , University of Massachusetts Lowell
  • Hongwei Zhu (external member), Professor, Operations and Information Systems Department, University of Massachusetts Lowell

Abstract:

Since the adoption of eXtensible Business Reporting Language (XBRL) in 2009 by the US Security and Exchange Commission, a large number of financial reports in XBRL format have been collected and made publicly available. However, this valuable and continually growing dataset is underutilized largely due to the lack of infrastructure and technology support. There is an imperative need for analyzing a broad range of data (of which XBRL data is a large part) to understand the complex financial and economic systems.

In this dissertation, we aim to develop an open knowledge infrastructure using knowledge graphs that will enable public access to financial data and support knowledge-based exploitation. With the completion of the financial knowledge graph, we study a challenging problem for knowledge graphs, querying and extracting potentially a large number of entities that are of a user’s interest. The conventional query mechanism of subgraph pattern matching would not work well as the user in general does not know the graph pattern to search for. Moreover, there may not be a single subgraph pattern that fits all the intended entities. Using keywords also may not be a viable approach, as it is very difficult to come up with the right set of keywords, and the results are often very diverse and overwhelming. We propose a novel approach that does not require users to know much about the knowledge graph but only simple keywords about the desired entities. We retrieve a sample of matches and learn the entity context patterns as what we call the subgraph sketch signatures. We provide clustered patterns for the user to prune. Moreover, we devise a novel index to finally perform efficient entity retrieval over the whole knowledge graph.

All interested students and faculty members are invited to attend the defense via remote access.