jacklanda/SemanticQA

SemanticQA是一个用于评估语言模型在语义短语处理上的综合基准数据集，源自论文《Revisiting a Pain in the Neck: Semantic Phrase Processing Benchmark for Language Models》。该数据集包含多个子集，涉及不同类型的语义短语任务，如搭配检索、搭配分类、搭配提取、搭配解释、习语检测、习语提取、习语解释、名词复合词构成性、名词复合词提取、名词复合词解释以及动词多词表达提取。数据集为英文单语，采用MIT许可证，适用于文本分类、文本生成和问答任务。

SemanticQA is a comprehensive benchmark for evaluating language models on semantic phrase processing, from the paper *Revisiting a Pain in the Neck: Semantic Phrase Processing Benchmark for Language Models*. It includes multiple subsets for various tasks related to semantic phrases such as collocate retrieval, collocation categorization, collocation extraction, collocation paraphrase, idiom detection, idiom extraction, idiom paraphrase, noun compound compositionality, noun compound extraction, noun compound interpretation, and verbal multiword expression extraction. The dataset is in English, monolingual, under the MIT license, and suitable for text-classification, text-generation, and question-answering tasks.