vgatm-main.zip (31.74 MB)
Download fileData and code for: Variational Graph Author Topic Modeling
dataset
posted on 2022-10-25, 04:48 authored by ZHANG, CE (SMU)ZHANG, CE (SMU), Hady Wirawan LAUWHady Wirawan LAUWThis is the tensorflow implementation of KDD-2022 paper "Variational Graph Author Topic Modeling" by Delvin Ce Zhang and Hady W. Lauw.
VGATM is a Graph Neural Network model that extracts interpretable topics for documents with authors and venues. Topics of documents then fulfill document classification, citation prediction, etc.