GeoWebNews
收藏资源简介:
我们还发布了一个新的GeoWebNews数据集,以挑战研究人员 开发机器学习 (ML) 算法来评估分类/标记 基于深层语用学而不是浅层句法特征的表现。 第2节给出了Geoparsing、NER、GIE和GIR的背景。我们呈现 第3节中的新分类法,描述和分类地名类型。在 第四节,我们对当前的评价方法进行了全面的回顾。 并证明推荐的框架是合理的。最后,第5节介绍了 GeoWebNews数据集、注释和资源。我们还评估地理标记 以及新数据集上的地名解析,说明了 几种序列标记模型,如SpacyNLP和谷歌NLP。
We also release a novel GeoWebNews dataset to challenge researchers to develop machine learning (ML) algorithms for evaluating the performance of classification/tagging tasks based on deep pragmatic rather than shallow syntactic features. Section 2 provides background on Geoparsing, NER, GIE, and GIR. We present a novel taxonomy in Section 3, which describes and classifies toponym types. In Section 4, we conduct a comprehensive review of current evaluation methods and justify the proposed framework. Finally, Section 5 introduces the GeoWebNews dataset, annotations, and resources. We also evaluate geotagging and geoparsing on the novel dataset, demonstrating several sequence tagging models such as spaCyNLP and Google NLP.




