Code for the paper "Query Embedding on Hyper-relational Knowledge Graphs"

Related tags

Deep LearningStarQE
Overview

Query Embedding on Hyper-Relational Knowledge Graphs

This repository contains the code used for the experiments in the paper

Query Embedding on Hyper-Relational Knowledge Graphs.
Dimitrios Alivanistos and Max Berrendorf and Michael Cochez and Mikhail Galkin

If you encounter any problems, or have suggestions on how to improve this code, open an issue.

Abstract: Multi-hop logical reasoning is an established problem in the field of representation learning on knowledge graphs (KGs). It subsumes both one-hop link prediction as well as other more complex types of logical queries. Existing algorithms operate only on classical, triple-based graphs, whereas modern KGs often employ a hyper-relational modeling paradigm. In this paradigm, typed edges may have several key-value pairs known as qualifiers that provide fine-grained context for facts. In queries, this context modifies the meaning of relations, and usually reduces the answer set. Hyper-relational queries are often observed in real-world KG applications, and existing approaches for approximate query answering cannot make use of qualifier pairs. In this work, we bridge this gap and extend the multi-hop reasoning problem to hyper-relational KGs allowing to tackle this new type of complex queries. Building upon recent advancements in Graph Neural Networks and query embedding techniques, we study how to embed and answer hyper-relational conjunctive queries. Besides that, we propose a method to answer such queries and demonstrate in our experiments that qualifiers improve query answering on a diverse set of query patterns.

Requirements

We developed our repository using Python 3.8.5. Other version may also work.

First, please ensure that you have properly installed

in your environment. Running experiments is possible on both CPU and GPU. On a GPU, the training should go noticeably faster. If you are using GPU, please make sure that the installed versions match your CUDA version.

We recommend the use of virtual environments, be it virtualenv or conda.

Now, clone the repository and install other dependencies using pip. After moving to the root of the repo (and with your virtual env activated) type:

pip install .

If you want to change code, we suggest to use the editable mode of the pip installation:

pip install -e .

To log results, we suggest using wandb. Instructions on installation and setting up can be found here: https://docs.wandb.ai/quickstart

Running test (optional)

You can run the tests by installing the test dependencies

pip install -e '.[test]'

and then executing them

pytest

Both from the root of the project.

It is normal that you see some skipped tests.

Running experiments

The easiest way to start experiments is via the command line interface. The command line also provides more information on the options available for each command. You can show the help it by typing

hqe --help

into a terminal within your active python environment. Some IDEs, e.g. PyCharm, require you to start from a file if you want to enable the debugger. To this end, we also provide a thin wrapper in executables, which you can start by

python executables/main.py

Downloading the data

To run experiments, we offer the preprocessed queries for download. It is also possible to run the preprocessing steps yourself, cf. the data preprocessing README, using the following command

hqe preprocess skip-and-download-binary

Training a model

There are many options are available for model training. For an overview of options, run

hqe train --help

Some examples:


Train with default settings, using 10000 reified 1hop queries with a qualifier and use 5000 reified triples from the validation set. Details on how to specify the amount of samples can be found in [src/mphrqe/data/loader.Sample](the Sample class). Note that the data loading is taking care of only using data from the correct data split.

hqe train \
    -tr /1hop/1qual:atmost10000:reify \
    -va /1hop/1qual:5000:reify

Train with the same data, but with custom parameters for the model. The example below uses target pooling to get the embedding of the query graph, uses a dropout of 0.5 in the layers, uses cosine similarity instead of the dot product to compute similarity when ranking answers to the query, and enables wandb for logging the metrics. Finally, the trained model is stored as a file training-example-model.pt which then be used in the evaluation.

hqe train \
    -tr /1hop/1qual:atmost10000:reify \
    -va /1hop/1qual:5000:reify \
    --graph-pooling TargetPooling \
    --dropout 0.5 \
    --similarity CosineSimilarity \
    --use-wandb --wandb-name "training-example" \
    --save \
    --model-path "training-example-model.pt"

By default, the model path is relative to the current working directory. Providing an absolute path to a different directory can change that.

Performing hyper parameter optimization

To find optimal parameters for a dataset, one can run a hyperparameter optimization. Under the hood this is using the optuna framework.

All options for the hyperparameter optimization can be seen with

hqe optimize --help

Some examples:


Run hyper-parameter optimization. This will result in a set of runs with different hyper-parameters from which the user can pick the best.

hqe optimize \
    -tr "/1hop/1qual-per-triple:*" \
    -tr "/2i/1qual-per-triple:atmost40000" \
    -tr "/2hop/1qual-per-triple:40000" \
    -tr "/3hop/1qual-per-triple:40000" \
    -tr "/3i/1qual-per-triple:40000" \
    -va "/1hop/1qual-per-triple:atmost3500" \
    -va "/2i/1qual-per-triple:atmost3500" \
    -va "/2hop/1qual-per-triple:atmost3500" \
    -va "/3hop/1qual-per-triple:atmost3500" \
    -va "/3i/1qual-per-triple:atmost3500" \
    --use-wandb \
    --wandb-name "hpo-query2box-style"

Evaluating model performance

To evaluate a model's performance on the test set, we provide an example below:

hqe evaluate \
    --test-data "/1hop/1qual:5000:reify" \
    --use-wandb \
    --wandb-name "test-example" \
    --model-path "training-example-model.pt"

Citation

If you find this work useful, please consider citing

@misc{alivanistos2021query,
      title={Query Embedding on Hyper-relational Knowledge Graphs}, 
      author={Dimitrios Alivanistos and Max Berrendorf and Michael Cochez and Mikhail Galkin},
      year={2021},
      eprint={2106.08166},
      archivePrefix={arXiv},
      primaryClass={cs.AI}
}
You might also like...
Code for the paper Learning the Predictability of the Future

Learning the Predictability of the Future Code from the paper Learning the Predictability of the Future. Website of the project in hyperfuture.cs.colu

PyTorch code for the paper: FeatMatch: Feature-Based Augmentation for Semi-Supervised Learning
PyTorch code for the paper: FeatMatch: Feature-Based Augmentation for Semi-Supervised Learning

FeatMatch: Feature-Based Augmentation for Semi-Supervised Learning This is the PyTorch implementation of our paper: FeatMatch: Feature-Based Augmentat

Code for the paper A Theoretical Analysis of the Repetition Problem in Text Generation
Code for the paper A Theoretical Analysis of the Repetition Problem in Text Generation

A Theoretical Analysis of the Repetition Problem in Text Generation This repository share the code for the paper "A Theoretical Analysis of the Repeti

Code for our ICASSP 2021 paper: SA-Net: Shuffle Attention for Deep Convolutional Neural Networks
Code for our ICASSP 2021 paper: SA-Net: Shuffle Attention for Deep Convolutional Neural Networks

SA-Net: Shuffle Attention for Deep Convolutional Neural Networks (paper) By Qing-Long Zhang and Yu-Bin Yang [State Key Laboratory for Novel Software T

Open source repository for the code accompanying the paper 'Non-Rigid Neural Radiance Fields Reconstruction and Novel View Synthesis of a Deforming Scene from Monocular Video'.
Open source repository for the code accompanying the paper 'Non-Rigid Neural Radiance Fields Reconstruction and Novel View Synthesis of a Deforming Scene from Monocular Video'.

Non-Rigid Neural Radiance Fields This is the official repository for the project "Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synt

Code for the Shortformer model, from the paper by Ofir Press, Noah A. Smith and Mike Lewis.

Shortformer This repository contains the code and the final checkpoint of the Shortformer model. This file explains how to run our experiments on the

PyTorch code for ICLR 2021 paper Unbiased Teacher for Semi-Supervised Object Detection
PyTorch code for ICLR 2021 paper Unbiased Teacher for Semi-Supervised Object Detection

Unbiased Teacher for Semi-Supervised Object Detection This is the PyTorch implementation of our paper: Unbiased Teacher for Semi-Supervised Object Detection

Official code for paper "Optimization for Oriented Object Detection via Representation Invariance Loss".

Optimization for Oriented Object Detection via Representation Invariance Loss By Qi Ming, Zhiqiang Zhou, Lingjuan Miao, Xue Yang, and Yunpeng Dong. Th

Code for our CVPR 2021 paper
Code for our CVPR 2021 paper "MetaCam+DSCE"

Joint Noise-Tolerant Learning and Meta Camera Shift Adaptation for Unsupervised Person Re-Identification (CVPR'21) Introduction Code for our CVPR 2021

Comments
  • bug in SPARQL for 1hop-2i/0qual

    bug in SPARQL for 1hop-2i/0qual

    It looks like the SPARQL is not executable. should line 37 in test/validation and line 22 in train: FILTER ((?s1 != ?o2_s0) || (?s1 = ?o2_s0 && str(?p0)< str(?1) )) be FILTER ((?s1 != ?o2_s0) || (?s1 = ?o2_s0 && str(?p0)< str(?p1) )) ?

    opened by Kelaproth 2
Releases(v1.0.0-iclr)
Owner
DimitrisAlivas
Researcher. Data scientist. Passionate about Tech & AI
DimitrisAlivas
A framework that constructs deep neural networks, autoencoders, logistic regressors, and linear networks

A framework that constructs deep neural networks, autoencoders, logistic regressors, and linear networks without the use of any outside machine learning libraries - all from scratch.

Kordel K. France 2 Nov 14, 2022
Composing methods for ML training efficiency

MosaicML Composer contains a library of methods, and ways to compose them together for more efficient ML training.

MosaicML 2.8k Jan 08, 2023
Self-Supervised Monocular DepthEstimation with Internal Feature Fusion(arXiv), BMVC2021

DIFFNet This repo is for Self-Supervised Monocular DepthEstimation with Internal Feature Fusion(arXiv), BMVC2021 A new backbone for self-supervised de

Hang 94 Dec 25, 2022
Invariant Causal Prediction for Block MDPs

MISA Abstract Generalization across environments is critical to the successful application of reinforcement learning algorithms to real-world challeng

Meta Research 41 Sep 17, 2022
Least Square Calibration for Peer Reviews

Least Square Calibration for Peer Reviews Requirements gurobipy - for solving convex programs GPy - for Bayesian baseline numpy pandas To generate p

Sigma <a href=[email protected]"> 1 Nov 01, 2021
DeOldify - A Deep Learning based project for colorizing and restoring old images (and video!)

DeOldify - A Deep Learning based project for colorizing and restoring old images (and video!)

Jason Antic 15.8k Jan 04, 2023
Image Matching Evaluation

Image Matching Evaluation (IME) IME provides to test any feature matching algorithm on datasets containing ground-truth homographies. Also, one can re

32 Nov 17, 2022
This is the pytorch re-implementation of the IterNorm

IterNorm-pytorch Pytorch reimplementation of the IterNorm methods, which is described in the following paper: Iterative Normalization: Beyond Standard

Lei Huang 32 Dec 27, 2022
Studying Python release adoptions by looking at PyPI downloads

Analysis of version adoptions on PyPI We get PyPI download statistics via Google's BigQuery using the pypinfo tool. Usage First you need to get an acc

Julien Palard 9 Nov 04, 2022
Implementation for ACProp ( Momentum centering and asynchronous update for adaptive gradient methdos, NeurIPS 2021)

This repository contains code to reproduce results for submission NeurIPS 2021, "Momentum Centering and Asynchronous Update for Adaptive Gradient Meth

Juntang Zhuang 15 Jun 11, 2022
a generic C++ library for image analysis

VIGRA Computer Vision Library Copyright 1998-2013 by Ullrich Koethe This file is part of the VIGRA computer vision library. You may use,

Ullrich Koethe 378 Dec 30, 2022
Texture mapping with variational auto-encoders

vae-textures This is an experiment with using variational autoencoders (VAEs) to perform mesh parameterization. This was also my first project using J

Alex Nichol 41 May 24, 2022
[ICCV'21] UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction

UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction Project Page | Paper | Supplementary | Video This reposit

331 Dec 28, 2022
an Evolutionary Algorithm assisted GAN

EvoGAN an Evolutionary Algorithm assisted GAN ckpts

3 Oct 09, 2022
Computer vision - fun segmentation experience using classic and deep tools :)

Computer_Vision_Segmentation_Fun Segmentation of Images and Video. Tools: pytorch Models: Classic model - GrabCut Deep model - Deeplabv3_resnet101 Flo

Mor Ventura 1 Dec 18, 2021
Joint-task Self-supervised Learning for Temporal Correspondence (NeurIPS 2019)

Joint-task Self-supervised Learning for Temporal Correspondence Project | Paper Overview Joint-task Self-supervised Learning for Temporal Corresponden

Sifei Liu 167 Dec 14, 2022
Tracking code for the winner of track 1 in the MMP-Tracking Challenge at ICCV 2021 Workshop.

Tracking Code for the winner of track1 in MMP-Trakcing challenge This repository contains our tracking code for the Multi-camera Multiple People Track

DamoCV 29 Nov 13, 2022
Fast, Attemptable Route Planner for Navigation in Known and Unknown Environments

FAR Planner uses a dynamically updated visibility graph for fast replanning. The planner models the environment with polygons and builds a global visi

Fan Yang 346 Dec 30, 2022
“英特尔创新大师杯”深度学习挑战赛 赛道3:CCKS2021中文NLP地址相关性任务

基于 bert4keras 的一个baseline 不作任何 数据trick 单模 线上 最高可到 0.7891 # 基础 版 train.py 0.7769 # transformer 各层 cls concat 明神的trick https://xv44586.git

孙永松 7 Dec 28, 2021
🔥 Cogitare - A Modern, Fast, and Modular Deep Learning and Machine Learning framework for Python

Cogitare is a Modern, Fast, and Modular Deep Learning and Machine Learning framework for Python. A friendly interface for beginners and a powerful too

Cogitare - Modern and Easy Deep Learning with Python 76 Sep 30, 2022