ekacare/indian_drug_mcqa
收藏资源简介:
该数据集是一个多项选择问答数据集,用于评估对印度品牌药物的知识,特别是从品牌名称识别通用名称或盐成分的能力。数据集包含1512个问题,涵盖20多个治疗类别的印度药品品牌名称。每个问题提供一个品牌药物,并要求从3-5个选项中选择正确的通用成分。此外,数据集还包括针对高难度药物的变体问题,使用不同的干扰策略和措辞风格以提高评估的鲁棒性。
This dataset is a multiple-choice question answering (MCQA) dataset dedicated to evaluating knowledge of Indian branded pharmaceuticals, specifically the ability to identify their generic names or salt constituents from their corresponding brand names. It comprises 1,512 questions covering Indian pharmaceutical brand names across more than 20 therapeutic categories. Each question presents a branded drug and requires respondents to select the correct generic ingredient from 3 to 5 options. Additionally, the dataset includes variant questions for high-difficulty medications, which adopt diverse distractor strategies and wording styles to enhance the robustness of the evaluation.




