relbert/semeval2012_relational_similarity

SemEval2012关系相似性数据集用于微调RelBERT模型，包含来自89个预定义关系的正负词对。这些关系基于10个父关系类型构建，每个父关系类型进一步细分为子关系类型。数据集的实例展示了关系类型、正例和负例的词对。数据集分为训练集和验证集，各包含79个实例。

The SemEval2012 Relation Similarity Dataset is designed for fine-tuning the RelBERT model. It consists of positive and negative word pairs across 89 predefined relations. These relations are built upon 10 parent relation types, with each parent type further subdivided into sub-relation types. The dataset instances include the relation type as well as the corresponding positive and negative word pairs. The dataset is split into a training set and a validation set, each containing 79 instances.

原始信息汇总

数据集概述

数据集名称

• 名称: SemEval2012 relational similarity dataset

数据集描述

• 来源: SemEval2012 task 2
• 用途: 用于fine-tune RelBERT模型
• 内容: 包含89种预定义关系下的正负词对列表

数据集结构

• 关系类型: 基于10种父关系类型构建，包括Class Inclusion, Part-Whole, Similar, Contrast, Attribute, Non Attribute, Case Relation, Cause-Purpose, Space-Time, Representation等。
• 数据实例: 每个实例包含关系类型、正向词对和负向词对。
• 数据分割: 训练集和验证集各包含79个实例。

引用信息

@inproceedings{jurgens-etal-2012-semeval, title = "{S}em{E}val-2012 Task 2: Measuring Degrees of Relational Similarity", author = "Jurgens, David and Mohammad, Saif and Turney, Peter and Holyoak, Keith", booktitle = "*{SEM} 2012: The First Joint Conference on Lexical and Computational Semantics {--} Volume 1: Proceedings of the main conference and the shared task, and Volume 2: Proceedings of the Sixth International Workshop on Semantic Evaluation ({S}em{E}val 2012)", month = "7-8 " # jun, year = "2012", address = "Montr{e}al, Canada", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/S12-1047", pages = "356--364", }