Mojtaba Nayyeri

Abstract

Modeling evolving interactions among entities is critical in many real-world tasks. For example, predicting driver maneuvers in traffic requires tracking how neighboring vehicles accelerate, brake, and change lanes relative to one another over consecutive frames. Likewise, detecting financial fraud hinges on following the flow of funds through successive transactions as they propagate through the network. Unlike classic time-series forecasting, these settings demand reasoning over who interacts with whom and when, calling for a temporal-graph representation that makes both the relations and their evolution explicit. Existing temporal-graph methods typically use snapshot graphs to encode temporal evolution. We introduce a full-history graph that instantiates one node for every entity at every time step and separates two edge sets: (i) intra-time-step edges that capture relations within a single frame and (ii) inter-time-step edges that connect an entity to itself at consecutive steps. To learn on this graph we design an Edge-Type Decoupled Network (ETDNet) with parallel modules: a graph-attention module aggregates information along intra-time-step edges, a multi-head temporal-attention module attends over an entity's inter-time-step history, and a fusion module combines the two messages after every layer. Evaluated on driver-intention prediction (Waymo) and Bitcoin fraud detection (Elliptic++), ETDNet consistently surpasses strong baselines, lifting Waymo joint accuracy to 75.6\% (vs. 74.1\%) and raising Elliptic++ illicit-class F1 to 88.1\% (vs. 60.4\%). These gains demonstrate the benefit of representing structural and temporal relations as distinct edges in a single graph.

Abstract

This is the official pytorch implementation of the paper "Faithful embeddings for EL++ Knowledge Bases" published in ISWC 2022. The code was implemented based on el-embeddings. The code can be used to reproduce the experiments on subsumption reasoning. To execute the code, follow the instructions in the README.md file. For more info, please check the paper. Please have no hesitation to contact the authors for any inquiries.

Abstract

This is a Pytorch implementation of the paper Shrinking Embeddings for Hyper-relational Knowledge Graphs published in ACL'23.This code is used to reproduce the experiments of the method ShrinkE, a geometric embedding approach for hyper-relational knowledge graphs. The code is implemented with Python 3 and pytorch. The code is tested on public datasets which can be download from StarE. To execute the code, follow the instructions in the README.md file. For more info, please check the paper or feel free to contact the authors for any inquiries.

Abstract

This code is a PyTorch implementation of the paper "NestE: Modeling Nested Relational Structures for Knowledge Graph Reasoning (AAAI'24)".NestE is a knowledge graph embedding method that can encode nested facts represented by quoted triples (h,r,t) in which the subject and object are triples themselves, e.g., ((BarackObama, holds_position, President), succeed_by, (DonaldTrump, holds_position, President)).We implement six variant models of NetsE based on different hypercomplex number systems. NestE_Q.py for NestE with quaternion. NestE_H.py for NestE with hyperbolic quaternion. NestE_D.py for NestE with split quaternion. NestE_B.py, NestE_HB.py, and NestE_DB.py are the respective version with a translation component. This code is used to reproduce the experiments of the paper. To execute the code, follow the instructions in the README.md file.

Abstract

This is a Pytorch implementation of the paper Ultrahyperbolic Knowledge Graph Embeddings published in KDD 2022. This code is used to reproduce the experiments of the method UltraE, a geometric embedding approach for knowledge graph embeddings. The code is tested on public datasets which can be downloaded from KGEmb. To execute the code, follow the instructions in the README.md file. For more info, please check the paper or feel free to contact the authors for any inquiries.

Abstract

This is a PyTorch implementation of the paper Hyperbolic Embedding Inference for Structured Multi-Label Prediction published in NeurIPS 2022. The code provides the Python scripts to reproduce the experiments in the paper, as well as a proof-of-concept example of the method. To execute the code, follow the instructions in the README.md file. For more info, please check the paper. Please have no hesitation to contact the authors for any inquiries.

Abstract

Temporal knowledge graphs represent temporal facts (s,p,o,τ) relating a subject s and an object o via a relation label p at time τ, where τ could be a time point or time interval. Temporal knowledge graphs may exhibit static temporal patterns at distinct points in time and dynamic temporal patterns between different timestamps. In order to learn a rich set of static and dynamic temporal patterns and apply them for inference, several embedding approaches have been suggested in the literature. However, as most of them resort to single underlying embedding spaces, their capability to model all kinds of temporal patterns was severely limited by having to adhere to the geometric property of their one embedding space. We lift this limitation by an embedding approach that maps temporal facts into a product space of several heterogeneous geometric subspaces with distinct geometric properties, i.e.\ Complex, Dual, and Split-complex spaces. In addition, we propose a temporal-geometric attention mechanism to integrate information from different geometric subspaces conveniently according to the captured relational and temporal information. Experimental results on standard temporal benchmark datasets favorably evaluate our approach against state-of-the-art models.