Search Results for author: Kin Sum Liu

Found 6 papers, 0 papers with code

An Analysis Of Entire Space Multi-Task Models For Post-Click Conversion Prediction

no code implementations18 Aug 2021 Conor O'Brien, Kin Sum Liu, James Neufeld, Rafael Barreto, Jonathan J Hunt

Industrial recommender systems are frequently tasked with approximating probabilities for multiple, often closely related, user actions.

Multi-Task Learning Recommendation Systems

Ricci-GNN: Defending Against Structural Attacks Through a Geometric Approach

no code implementations1 Jan 2021 Ze Ye, Tengfei Ma, Chien-Chun Ni, Kin Sum Liu, Jie Gao, Chao Chen

We propose a novel GNN defense algorithm against structural attacks that maliciously modify graph topology.

Layered Graph Embedding for Entity Recommendation using Wikipedia in the Yahoo! Knowledge Graph

no code implementations15 Apr 2020 Chien-Chun Ni, Kin Sum Liu, Nicolas Torzec

In this paper, we describe an embedding-based entity recommendation framework for Wikipedia that organizes Wikipedia into a collection of graphs layered on top of each other, learns complementary entity representations from their topology and content, and combines them with a lightweight learning-to-rank approach to recommend related entities on Wikipedia.

Graph Embedding Learning-To-Rank

Topology Based Scalable Graph Kernels

no code implementations15 Jul 2019 Kin Sum Liu, Chien-Chun Ni, Yu-Yao Lin, Jie Gao

We propose a new graph kernel for graph classification and comparison using Ollivier Ricci curvature.

Graph Classification

Performing Co-Membership Attacks Against Deep Generative Models

no code implementations24 May 2018 Kin Sum Liu, Chaowei Xiao, Bo Li, Jie Gao

We conduct extensive experiments on a variety of datasets and generative models showing that: our attacker network outperforms prior membership attacks; co-membership attacks can be substantially more powerful than single attacks; and VAEs are more susceptible to membership attacks compared to GANs.

Cannot find the paper you are looking for? You can Submit a new open access paper.