Multiple Purpose Annotation using SLAT — Segment and Link-based Annotation Tool —
Masaki Noguchi, Kenta MIYOSHI, Takenobu Tokunaga, Ryu Iida, Mamoru Komachi, Kentaro Inui · Linguistic Annotation Workshop · 2008
In recent years, the use of large scale corpora in NLP applications, such as statistical parsing, has become prominent. As their use gained credibility, naturally so did the types of information they provided. There exist today many groups that create corpora: ANC, SFB at the University of Potsdam, just to name a few. In many cases these groups also provide specialized annotation tools for their corpora. However, these tools are just that: specialized, i.e. designed to work with a very specific annotation definition, without flexibility in mind. In the early stages of a project, often times the specification for annotating changes. This makes it difficult to use a tool with such rigid boundaries. In this paper, we propose a browser-based annotation tool SLAT, which allows for easily adding and customizing annotations. We also explain the steps involved in customizing SLAT to meet a user’s project needs.