CaLiGraph for Semantic Reasoning Evaluation Challenge

This repository contains code and data to use CaLiGraph as a benchmark dataset in the Semantic Reasoning Evaluation Challange at the International Semantic Web Conference 2021 (ISWC'21).

The paper describing the dataset characteristics and results for well-known reasoners will be linked here as soon as a preprint is available.

Datasets

We use CaLiGraph version 2.1.0 as foundation for the challenge dataset. In particular, we use the files caligraph-ontology.nt.bz2 and caligraph-instances_types.nt.bz2 to generate our sample data.

We provide sample datasets having roughly 10ⁿ classes with n in [1,6]. The datasets and all potentially inferrable assertions can be found here. Please refer to our paper if you are interested in how the sample datasets were constructed.

Evaluated Reasoners

We evaluated the following reasoners with our sample datasets:

To evaluate the reasoners, we used their connectors in OWL API. The source code for the evaluation of the reasoners can be found in the folder reasoner_evaluation. As Pellet needs an OWL API version lower than 4 (while the others need a version higher than 4) we provide two pom.xml files. Depending on which reasoner you want to run, you have to use the correct one. Further, you have to uncomment the respective reasoners in the getReasoners() function of the java file org.unima.nheist.App.

Alternatively, you can use the two provided jar files to run the reasoners with the datasets. First download the sample dataset you want to run the reasoners on, then run the jar file and provide the location of the dataset as first argument like this:

java -jar semrec-caligraph-elk-hermit.jar <PATH-TO-DATASET-FILE> &> log_elk-hermit.txt

The result is a realization of the input dataset through the selected reasoners. Have a look at the log file (in the case of the example: log_elk-hermit.txt) for additional information about the reasoning process.

The precomputed results for all the sample datasets can be found here.

CaLiGraph Ontology as a Challenge for Semantic Reasoners ([email protected]'21)

Related tags

Overview

CaLiGraph for Semantic Reasoning Evaluation Challenge

Datasets

Evaluated Reasoners

Owner

Nico Heist

Generating Band-Limited Adversarial Surfaces Using Neural Networks

Wenet STT Python

Contrastive Learning for Metagenomic Binning

AI drive app that can help user become beautiful.

Interpretable and Generalizable Person Re-Identification with Query-Adaptive Convolution and Temporal Lifting

Noether Networks: meta-learning useful conserved quantities

Learning Time-Critical Responses for Interactive Character Control

Implementation of U-Net and SegNet for building segmentation

LightLog is an open source deep learning based lightweight log analysis tool for log anomaly detection.

This is an official pytorch implementation of Fast Fourier Convolution.

Arch-Net: Model Distillation for Architecture Agnostic Model Deployment

OpenL3: Open-source deep audio and image embeddings

OoD Minimum Anomaly Score GAN - Code for the Paper 'OMASGAN: Out-of-Distribution Minimum Anomaly Score GAN for Sample Generation on the Boundary'

Project to create an open-source 6 DoF input device

Code base for the paper "Scalable One-Pass Optimisation of High-Dimensional Weight-Update Hyperparameters by Implicit Differentiation"

Array Camera Ptychography

Source code and dataset of the paper "Contrastive Adaptive Propagation Graph Neural Networks forEfficient Graph Learning"

2.86% and 15.85% on CIFAR-10 and CIFAR-100

MQBench: Towards Reproducible and Deployable Model Quantization Benchmark

Video Corpus Moment Retrieval with Contrastive Learning (SIGIR 2021)