Lifelong sequential modeling for user response prediction. A comprehensive evaluation framework for our SIGIR 2019 paper.

User-Centric Conversational Recommendation with Multi-Aspect User Modeling (UCCR)

[WSDM 2024 Oral] This is our Pytorch implementation for the paper: "Intent Contrastive Learning with Cross Subsequences for Sequential Recommendation".

AMBAL-based NILM Trace generator

SCoRe is a sequential recommendation model with dual side neighbor-based collaborative filtering. Implementation of our WSDM 2020 paper.

The official PyTorch implementation of "Learning to Simulate Daily Activities via Modeling Dynamic Human Needs" (WWW'23)

Implementation of "Knowing your FATE: Friendship, Action and Temporal Explanations for User Engagement Prediction on Social Apps"

Our CIKM21 Paper "Incorporating Query Reformulating Behavior into Web Search Evaluation"

Predicting Anchor Links between Heterogeneous Social Networks (ASONAM 2016)

UserBERT PyTorch implementation

Repository of the paper "Toward a Responsible Fairness Analysis: From Binary to Multiclass and Multigroup Assessment in Graph Neural Network-Based User Modeling Tasks"

Dynamic Collaborative Filtering powered by Matrix and Tensor Integrators

Bachelor Thesis on Dynamic User Preferences in Recommendation Systems using Deep Reinforcement Learning

Python and Neo4J used to create a property graph database for ML algorithms - MS Thesis

The AMBAL-based NILM Trace generator (for NILMTK)

On Anonymous Commenting: A Greedy Approach to Balance Utilization and Anonymity for Instagram Users - Accepted at SIGIR 2019