Topic Modeling on Document Networks with Adjacent-Encoder
Speaker (s):
ZHANG Ce
PhD Candidate
School of Information Systems
Singapore Management University
|
|
Date:
Time:
Venue:
|
|
22 January 2020, Wednesday
2:30pm - 3:00pm
Meeting Room 4.4, Level 4
School of Information Systems
Singapore Management University
80 Stamford Road
Singapore 178902
We look forward to seeing you at this research seminar.
|
|
About the Talk
Oftentimes documents are linked to one another in a network structure, e.g., academic papers cite other papers, Web pages link to other pages. In this paper, we propose a holistic topic model to learn meaningful and unified low-dimensional topic representations for networked documents that seek to preserve both textual content and network structure. On the basis of reconstructing not only the input document but also its adjacent neighbors, we develop two neural encoder architectures. Adjacent-Encoder induces competition among documents for topic propagation, and reconstruction among neighbors for semantic capture. Adjacent-Encoder-X extends this to also encode the network structure in addition to document content.
This is a pre-conference talk for Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI-20).
About the Speaker
ZHANG Ce is a PhD student in Computer Science at the SMU School of Information Systems. He received his B.Eng in Computer Science and B.Econ in Financial Engineering from Sichuan University, P.R.China, both with outstanding graduate award. Together with his supervisor Prof. Hady W. Lauw at SMU, he is currently researching graph representation learning on homogeneous and heterogeneous networks, as well as their applications, such as recommender systems and data visualization.
