Knowledge Graph Parser
Overview
kg-parser is a Python package and CLI application that extracts structured knowledge graph triples from unstructured text using large language models (LLMs). The package supports multiple backends (HuggingFace, OpenAI, and local Jan) and offers flexible output formats—either as arrays of strings or as dictionaries.
Features
- Multi-backend Support: Use HuggingFace, OpenAI, or a local Jan server.
- Batch Processing: Process multiple texts efficiently.
- Flexible Output: Choose between list or dict representations for triples.
- CLI & API: Easily run as a command-line tool or integrate into your Python projects.
Installation
Using Conda
Create and activate the conda environment with the provided configuration:
conda env create -f environment.yml
conda activate kg-parser
Using Pip
Install the package directly from PyPI:
pip install kg-parser
Install the minimal dependencies from requirements.txt:
pip install -r requirements.txt
Alternatively, install the package in editable mode:
pip install -e .
Usage
As a CLI Application
Run the CLI tool from the command line:
python -m kg_parser.cli \
--model-type huggingface \
--model-name-or-path "google/flan-t5-small" \
--input-file test_input.json \
--output-file output_kg.json \
--triple-format list
Arguments:
--model-type: Choose fromhuggingface,openai, orjan_local.--model-name-or-path: Specify the model name or path.--input-file: Path to a JSON file containing an array of text strings.--output-file: Path where the output JSON will be saved.--triple-format: Output format for triples (listfor arrays ordictfor dictionaries).
As a Python Package
Import and use kg-parser in your own Python scripts:
from kg_parser.config import ModelConfig, ModelType
from kg_parser.core import KGParser
# Configure the model
model_config = ModelConfig(
model_type=ModelType.HUGGINGFACE,
model_name_or_path="google/flan-t5-small"
)
parser = KGParser(model_config)
# Process texts
texts = [
"Mount Everest is the highest mountain in the world. It's located in Nepal."
]
results = parser.parse_batch(texts)
# Save results with triples as lists
parser.save_to_json(results, "output_kg.json", triple_format="list")
License
This project is licensed under the MIT License. See the LICENSE file for details.
Acknowledgements
- The knowledge graph extraction prompt is adapted from the paper GraphEval: A Knowledge-Graph Based LLM Hallucination Evaluation Framework. The original unmodified prompt can be found in Appendix A ("A. KG Construction Prompt").
Metadata
Release files for kg-parser 0.1.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| kg_parser-0.1.1.tar.gz | 11.1 kB | Without provenance | Details for kg_parser-0.1.1.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| kg_parser-0.1.1-py3-none-any.whl | Python 3 | none | any | Details for kg_parser-0.1.1-py3-none-any.whl |
Total release size: 22.5 kB
Release files / kg_parser-0.1.1.tar.gz
| Download URL | kg_parser-0.1.1.tar.gz |
|---|---|
| Size | 11.1 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
1a178fbf5ac5febc00c24daf7e3625f5f725bce0475aadab9c24d7ed523db873
|
|
BLAKE2b-256 checksum How to use checksums |
3a46ee4fbdbcbe4afc694ee6b5408c6764d92a2ff99d6c445ea2903283793fe9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.10.16
|
Release files / kg_parser-0.1.1-py3-none-any.whl
| Download URL | kg_parser-0.1.1-py3-none-any.whl |
|---|---|
| Size | 11.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
927584b40906718dc0abedb38ad29aa1d4d6d7aa9031506f263060706b13be7b
|
|
BLAKE2b-256 checksum How to use checksums |
83d55f882ee34e599811210fd1f8c87713e40c228655dd176d0d2726134042d8
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.10.16
|
Security for kg-parser 0.1.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.