datalyes/effect2substance

Effect to Substance Retrieval

Dataset Description

This dataset is part of PatenTEB, a comprehensive benchmark for evaluating text embedding models on patent-specific tasks. PatenTEB comprises 15 tasks across retrieval, classification, paraphrase detection, and clustering, with 2.06 million examples designed to reflect real-world patent analysis workflows.

Paper: PatenTEB: A Comprehensive Benchmark and Model Family for Patent Text Embedding

Task Details

• Task Name: effect2substance
• Task Type: Retrieval
• Test Samples: 2,037

Asymmetric retrieval task connecting described effects to the materials/substances achieving them. Evaluates whether models can bridge role-complementary relationships for analogical reasoning in patent examination.

Dataset Structure

This is a retrieval task where models find relevant patents given a query.

Splits:

• test: Query-document pairs for retrieval evaluation

Columns:

• q
• effect
• substance

Data Sample

Below is a 5-row preview of the test set:

q,effect,substance
048-502-794-320-147,"invention provides a drug delivery system in which the drug can be delivered directly to an affected site by means of an injection device, such as ...","invention refers to a derivative of photoreactive hyaluronic acid with an administered drug which contains a photoreactive group and a drug, and al..."
174-233-067-598-578,novel compounds have hepatitis c virus replication inhibitory action and can be used in medicine,"invention describes novel macrocyclic compounds of formulae pharmaceutically acceptable salts or stereoisomers thereof, where r1 = -or5, -nh-so2r6;..."
168-937-692-297-533,"obtaining a novel compound and its salt, which can be used in medicine as a medicinal agent for preventing or treating diseases such as nutritional...","invention relates to organic chemistry, and specifically to (2r)-2-{(1s)-6,7-dimethoxy-1-[2-(4-trifluoromethylphenyl)-ethyl]-3,4-dihydro-1h-isoquin..."
013-011-666-440-682,"improved reduction from sulfur poisoning, prevention of damage and failure of catalyst filter",group of inventions may be used to clean exhaust gas of internal combustion engines. device for cleaning of exhaust gas comprises device for supply...
106-994-082-017-305,simplified attachment and detachment of the auxiliary tools without fluid removal from the electric drive,"electric drive has electric terminals for attaching auxiliary tools, and it can include a hydraulic connector which ensures fluid containment in th..."

Evaluation Metrics

This task uses NDCG@10 (Normalized Discounted Cumulative Gain at rank 10) as the primary metric. NDCG measures ranking quality by discounting relevance scores by logarithmic position, normalized by the ideal ranking.

Usage

Load Dataset

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("datalyes/{task_name}")

# Access test split
test_data = dataset['test']

Use with Sentence Transformers

from sentence_transformers import SentenceTransformer

# Load a patent-specialized model
model = SentenceTransformer("datalyes/patembed-base")

# Encode patent texts
embeddings = model.encode(test_data['text'])

Integrate with MTEB

This dataset is designed to be integrated with the MTEB (Massive Text Embedding Benchmark) framework. Integration with MTEB is in progress and will be available once the corresponding pull requests are accepted.

Benchmark Context

This dataset is part of a larger benchmark suite:

Benchmark Component	Description
PatenTEB	15 tasks covering retrieval, classification, paraphrase, clustering
Test Data (Released)	319,320 examples across all 15 tasks
Training/Validation Data	1.74 million examples (planned for future release)
Total Dataset Size	2.06 million annotated instances

Note: Currently, only the test split is publicly available. Training and validation data release is planned for a future date.

All 15 Tasks (NEW to MTEB):

• 3 classification tasks: Bloom timing, NLI directionality, IPC3 classification
• 2 clustering tasks: IPC-based, Inventor-based
• 8 retrieval tasks: 3 symmetric (IN/MIXED/OUT domain) + 5 asymmetric (fragment-to-full)
• 2 paraphrase tasks: Problem and solution paraphrase detection

MTEB Integration: Upcoming (PR in progress)

Citation

If you use this dataset, please cite our paper:

@misc{ayaou2025patentebcomprehensivebenchmarkmodel,
      title={PatenTEB: A Comprehensive Benchmark and Model Family for Patent Text Embedding}, 
      author={Iliass Ayaou and Denis Cavallucci},
      year={2025},
      eprint={2510.22264},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2510.22264}
}

License

This dataset is released under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.

• You are free to share and adapt the material
• You must give appropriate credit
• You may not use the material for commercial purposes
• If you remix, transform, or build upon the material, you must distribute your contributions under the same license

For full license details, see: https://creativecommons.org/licenses/by-nc-sa/4.0/

Contact

• Authors: Iliass Ayaou, Denis Cavallucci
• Institution: ICUBE Laboratory, INSA Strasbourg
• GitHub: github.com/iliass-y/patenteb
• HuggingFace: huggingface.co/datalyes