Annotating Geographical Entities on Microblog Text

Abstract

This paper presents a discussion of the problems surrounding the task of annotating geographical entities on microblogs and reports the preliminary results of our efforts to annotate Japanese microblog texts. Unlike prior work, we not only annotate geographical location entities but also facility entities, such as stations, restaurants, shopping stores, hospitals and schools. We discuss ways in which to build a gazetteer, the types of ambiguities that need to be considered, reasons why the annotator tends to disagree, and the problems that need to be solved to automate the task of annotating the geographical entities. All the annotation data and the annotation guidelines are publicly available for research purposes from our web site.

UNH at SemEval-2019 Task 12: Toponym Resolution in Scientific Papers

“…There is significant work in the area of toponym detection (Matsuda et al, 2015;D. Lieberman et al, 2010) and the closely related fields of named entity recognition (NER) (Li et al, 2018) and entity mention detection (EMD) (Shen et al, 2015) with many different approaches.…”

Proceedings of the Workshop on Extracting Structured Knowledge from Scientific Publications

“…There is significant work in the area of toponym detection (Matsuda et al, 2015;Lieberman et al, 2010) and the closely related fields of named entity recognition and entity mention detection (Shen et al, 2015) with many different approaches. State-of-the-art named entity detection models have historically employed a combination of hand-crafted features, rules, natural language processing, string-pattern matching, and domain knowledge using supervised learning on relatively-small manually annotated corpora (Piskorski and Yangarber, 2013).…”

An Analysis of Deep Contextual Word Embeddings and Neural Architectures for Toponym Mention Detection in Scientific Publications

“…There is significant work in the area of toponym detection (Matsuda et al, 2015;Lieberman et al, 2010) and the closely related fields of named entity recognition (Li et al, 2018) and entity mention detection (Shen et al, 2015) with many different approaches. State-of-the-art named entity detection models have historically employed a combination of hand-crafted features, rules, natural language processing, string-pattern matching, and domain knowledge using supervised learning on relatively-small manually annotated corpora (Piskorski and Yangarber, 2013).…”