Multiperspective/multip_human_soft

--- dataset_info: features: - name: QID_x dtype: int64 - name: Topic dtype: string - name: Query dtype: string - name: docID dtype: string - name: docURL dtype: string - name: docTitle dtype: string - name: docCont dtype: string - name: engineID dtype: int64 - name: rank dtype: int64 - name: answer1 dtype: string - name: answer2 dtype: string - name: answer3 dtype: string - name: If there is a link-broken option dtype: int64 - name: majority_label dtype: string - name: labels dtype: string - name: Input dtype: string - name: Input_length dtype: int64 - name: docCont_length dtype: int64 - name: gpt_summaries dtype: string - name: doc dtype: string - name: label dtype: string - name: label_indices dtype: string - name: soft_labels dtype: string splits: - name: train num_bytes: 11133322 num_examples: 734 - name: test num_bytes: 2330101 num_examples: 157 - name: validation num_bytes: 2273070 num_examples: 157 download_size: 9426795 dataset_size: 15736493 configs: - config_name: default data_files: - split: train path: data/train-* - split: test path: data/test-* - split: validation path: data/validation-* ---

搜集汇总

背景与挑战

背景概述

该数据集是一个多视角人类软标签数据集，包含查询与文档对，以及人工标注的多个答案和软标签，用于文本相关性或问答系统的训练与评估。数据集分为训练、测试和验证集，共约1048个样本。