Exploring Sensorial Features for Metaphor Identification

Abstract

Language is the main communication device to represent the environment and share a common understanding of the world that we perceive through our sensory organs. Therefore, each language might contain a great amount of sensorial elements to express the perceptions both in literal and figurative usage. To tackle the semantics of figurative language, several conceptual properties such as concreteness or imegeability are utilized. However, there is no attempt in the literature to analyze and benefit from the sensorial elements for figurative language processing. In this paper, we investigate the impact of sensorial features on metaphor identification. We utilize an existing lexicon associating English words to sensorial modalities and propose a novel technique to automatically discover these associations from a dependency-parsed corpus. In our experiments, we measure the contribution of the sensorial features to the metaphor identification task with respect to a state of the art model. The results demonstrate that sensorial features yield better performance and show good generalization properties.

Sensory language across lexical categories

“…Dataset 3 is the Sensicon (Tekiroğlu et al, 2014), a resource which includes 22,684 lexemes together with their degree of association with the five senses. A second list has subsequently been published by the same authors (Tekiroğlu et al, 2015), however, this list does not include verbs and thus cannot be used to address the present research hypotheses. The Sensicon dataset we use here was constructed by looking at word co-occurrence statistics in the GigaWord corpus.…”

“…The Sensicon dataset we use here was constructed by looking at word co-occurrence statistics in the GigaWord corpus. The idea here is that if a given word occurs very frequently together in the same text with a particular seed word of a given sensory modality, then it is a word of this sensory modality (for problems with this assumption, see Louwerse & Connell, 2011, Winter, 2016b, and Tekiroğlu et al, 2015. The structure of the dataset is similar to the ratings represented in Connell (2009, 2013) and Winter (2016a), with a continuous numerical value for each modality association.…”

Developing New Software to Analyze the Emo-‎Sensory Load of Language

“…Understanding the world through sensory information coming from our sensory organs has been a significant issue for a long time (Tekiroğlu et al, 2014). Senses are classified into five categories dating back to Aristotle, namely sight, hearing, smell, taste, and touch, and they can be considered human beings' information channels that connect us to the outside environment.…”

“…To have a dataset in which the association degree of words can be determined, Tekiroğlu et al (2014) adopted a computational approach including two phases; in the first phase, they generated a sufficient number of sensory seed words. They employed the bootstrapping strategy and generated these numbers of sensory seed words from a small set of the manually selected seed of sensory words.…”

Recent advances in neural metaphor processing: A linguistic, cognitive and social perspective

“…It therefore makes sense to employ categorial features for metaphor identification. Tekiroglu et al (2015) tested the use of sensorial categories (the five human senses) for identifying AN synaes-thetic metaphors (e.g., sweet music, soft light). Using sensorial categories in addition to Word-Net supersenses, concreteness, and imageability led to improved performance (accuracy 0.890 vs 0.845 on TSV).…”