FUH (FernUniversität in Hagen): Metonymy Recognition Using Different Kinds of Context for a Memory-Based Learner

Proceedings of the 4th International Workshop on Semantic Evaluations (SemEval-2007), pages 153–156, Prague, June 2007. c⃝2007 Association for Computational Linguistics FUH (FernUniversit¨at in Hagen): Metonymy Recognition Using Different Kinds of Context for a Memory-Based Learner Johannes Leveling Intelligent Information and Communication Systems (IICS) FernUniversit¨at in Hagen (University of Hagen) johannes.leveling@fernuni-hagen.de Abstract For the metonymy resolution task SemEval-2007, the use of a memory-basedat learner to train classifiers for the identifica- tion of metonymic location names is inves- tigated. Metonymy is resolved on different levels of granularity, differentiating between literal and non-literal readings on the coarse level; literal, metonymic, and mixed read- ings on the medium level; and a number of classes covering regular cases of metonymy on a fine level. Different kinds of context are employed to obtain different features: 1) a sequence of n1 synset IDs represent- ing subordination information for nouns and for verbs, 2) n2 prepositions, articles, modal, and main verbs in the same sentence, and 3) properties of n3 tokens in a context window to the left and to the right of the location name. Different classifiers were trained on the Mascara data set to determine which values for the context sizes n1, n2, and n3 yield the highest accuracy (n1 = 4, n2 = 3, and n3 = 7, determined with the leave-one- out method). Results from these classifiers served as features for a combined classifier. In the training phase, the combined classifier achieved a considerably higher precision for the Mascara data. In the SemEval submis- sion, an accuracy of 79.8% on the coarse, 79.5% on the medium, and 78.5% on the fine level is achieved (the baseline accuracy is 79.4%). 1 Introduction Metonymy is typically defined as a figure of speech in which a speaker uses one entity to refer to an- other that is related to it (Lakoff and Johnson, 1980). The identification of metonymy becomes important for NLP tasks such as question answering (Stallard, 1993) or geographic information retrieval (Leveling and Hartrumpf, 2006). For regular cases of metonymy for locations and organizations, Markert and Nissim have proposed a set of metonymy classes. Annotating a subset of the BNC (British National Corpus), they extracted a set of metonymic proper nouns from two categories: country names (Markert and Nissim, 2002) and or- ganization names (Nissim and Markert, 2003). In the metonymy resolution task at SemEval- 2007, the goal was to identify metonymic names in a subset of the BNC. The task consists of two subtasks for company and country names, which are further divided into classification on a coarse level (recog- nizing literal and non-literal readings), on a medium level (differentiating non-literal readings into mixed and metonymic readings), and on a fine level (iden- tifying classes of regular metonymy, such as a name referring to the population, place-for-people). The task is described in more detail by Markert and Nis- sim (2007). 2 System Description 2.1 Tools and Resources The following tools and resources are used for the metonymy classification: • TiMBL 5.1 (Daelemans et al., …

Content truncated to 500 words