Dataset Card for Nayose-Bench-Instruction

This dataset was created as a benchmark for the entity resolution task in the pharmaceutical domain.

Dataset Details

This dataset is designed for the entity resolution task in the pharmaceutical domain. The entity resolution task refers to a paraphrasing task, such as rephrasing drug names, converting chemical substances into brand names, or rewriting chemical substances into chemical formulas.

Uses

from datasets import load_dataset
load_dataset("EQUES/Nayose-Bench-Instruction")

{
  "id": 9131,
  "instruction": "チロソールの異なる表記は?",
  "response": "チロソールの異なる表記は4-ヒドロキシフェニルエタノールです。",
  "task": "replace",
  "name_q": "チロソール",
  "name_a": "4-ヒドロキシフェニルエタノール"
}

Dataset Structure

The data is stored in a JSONL file, where each record consists of the following fields: "id", "instruction", "response", "task", "name_q", "name_a".

Example:

{"id": 159, "instrcution": "イソアルボリノールをを化学構造で書くと?", "response": "イソアルボリノールをを化学構造で書くとC30H50Oです。", "task": "formula", "name_q": "イソアルボリノール", "name_a": "C30H50O"}

Source Data

This dataset was created by processing the KEGG DRUG Database, a database that centrally aggregates pharmaceutical information from the perspective of chemical structures and components.

Dataset Card Contact

Created by Takuro Fujii ([email protected])

Acknowledgement

本データセットは、経済産業省及び国立研究開発法人新エネルギー・産業技術総合開発機構（NEDO）による生成AI開発力強化プロジェクト「GENIAC」により支援を受けた成果の一部である。