Proceedings of the First Workshop on Metaphor in NLP, Atlanta, Georgia, 13 June 2013
The First Workshop on Metaphor in NLP is the first one focused on modelling of metaphor using NLP techniques. The selected papers offer explorations into the following directions: (1) creation of metaphor-annotated datasets; (2) identification of new features that are useful for metaphor identification; (3) cross-lingual metaphor identification.The papers represent a variety of approaches to utilization and creation of datasets. While existing annotated corpora were used in some papers (Dunn, Tsvetkov et al), most papers describe creation of new annotated materials. Along with annotation guidelines adapted from the MIP and MIPVU procedures (Badryzlova et al), more intuitive annotation protocols are explored in Beigman Klebanov and Flor, Hovy et al, Heintz et al, Mohler et al, and Strzalkowski et al.
This work presents the tentative version of the protocol designed for annotation of a Russian metaphor corpus using the rapid annotation tool BRAT. The first part of the article is devoted to the procedure of "shallow" annotation in which metaphor-related words are identified according to a slightly modified version of the MIPVU procedure. The paper presents the results of two reliability tests and the measures of inter-annotator agreement obtained in them. Further on, the article gives a brief account of the linguistic problems that were encountered in adapting MIPVU to Russian. The rest of the first part describes the classes of metaphor-related words and the rules of their annotation with BRAT. The examples of annotation show how the visualization functionalities of BRAT allow the researcher to describe the multifaceted nature of metaphor related words and the complexity of their relations. The second part of the paper speaks about the annotation of conceptual metaphors (the "deep" annotation), where formulations of conceptual metaphors are inferred from the basic and contextual meanings of metaphor-related words from the "shallow" annotation, which is expected to make the metaphor formulation process more controllable.