Open-source knowgl-large model - Generate associated knowledge from text, extract entity relationships, and link to Wikidata

Knowgl Large

Developed by ibm-research

KnowGL is a model for generating and linking knowledge from text, capable of extracting entity-relation triples and linking them to Wikidata.

Knowledge Graph

Transformers

English#Knowledge Triple Generation #Wikidata Linking #State-of-the-art Relation Extraction

Downloads 247

Release Time : 9/20/2022

Model Overview

KnowGL is trained by combining Wikidata with extended training data from the REBEL dataset, enabling it to generate formatted triple information from given sentences and map labels and their types to Wikidata IDs.

Model Features

Knowledge Generation and Linking

Capable of generating structured triples from text and linking entities and relations to Wikidata.

Multi-task Processing

Simultaneously performs multiple tasks such as relation extraction, entity linking, and type linking.

State-of-the-art Performance

Achieves state-of-the-art relation extraction results on the REBEL dataset.

Model Capabilities

Text Understanding

Relation Extraction

Entity Recognition

Knowledge Graph Construction

Information Linking

Use Cases

Knowledge Management

Knowledge Graph Construction

Extracts structured knowledge from unstructured text to build a knowledge graph.

Automatically generates formatted triple information.

Information Extraction

Scientific Literature Analysis

Extracts entity-relation information from scientific literature.

Identifies research entities and their relationships.

🚀 KnowGL: Knowledge Generation and Linking from Text

KnowGL is a model that combines Wikidata with extended training data from the REBEL dataset. It can generate triples from text and achieve state - of - the - art results in relation extraction.

🚀 Quick Start

The knowgl-large model is trained by combining Wikidata with an extended version of the training data in the REBEL dataset. Given a sentence, KnowGL generates triple(s) in the following format:

[(subject mention # subject label # subject type) | relation label | (object mention # object label # object type)]

If there are more than one triples generated, they are separated by $ in the output. More details in Rossiello et al. (AAAI 2023).

The model achieves state - of - the - art results for relation extraction on the REBEL dataset. See results in Mihindukulasooriya et al. (ISWC 2022).

The generated labels (for the subject, relation, and object) and their types can be directly mapped to Wikidata IDs associated with them.

📚 Documentation

Model Information

Property	Details
Model Type	Seq2Seq
Training Data	Wikidata combined with an extended version of the training data in the REBEL dataset
Task Types	Relation - Extraction, Triple - Generation, Entity - Linking, Entity - Type - Linking, Relation - Linking
Datasets	Babelscape/rebel - dataset

Results

Task	Dataset	Metrics	Value
Relation Extraction	Babelscape/rebel - dataset (REBEL)	RE+ Macro F1	70.74

📄 License

This project is licensed under the CC - BY - NC - SA 4.0 license.

📚 Citation

@inproceedings{DBLP:conf/aaai/RossielloCMCG23,
  author       = {Gaetano Rossiello and
                  Md. Faisal Mahbub Chowdhury and
                  Nandana Mihindukulasooriya and
                  Owen Cornec and
                  Alfio Massimiliano Gliozzo},
  title        = {KnowGL: Knowledge Generation and Linking from Text},
  booktitle    = {{AAAI}},
  pages        = {16476--16478},
  publisher    = {{AAAI} Press},
  year         = {2023}
}

@inproceedings{DBLP:conf/semweb/Mihindukulasooriya22,
  author    = {Nandana Mihindukulasooriya and
               Mike Sava and
               Gaetano Rossiello and
               Md. Faisal Mahbub Chowdhury and
               Irene Yachbes and
               Aditya Gidh and
               Jillian Duckwitz and
               Kovit Nisar and
               Michael Santos and
               Alfio Gliozzo},
  title     = {Knowledge Graph Induction Enabling Recommending and Trend Analysis:
               {A} Corporate Research Community Use Case},
  booktitle = {{ISWC}},
  series    = {Lecture Notes in Computer Science},
  volume    = {13489},
  pages     = {827--844},
  publisher = {Springer},
  year      = {2022}
}

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご