We describe an unusual data set of thousands of annotated images with interesting sense phenomena. Natural language image sense annotation involves increased semantic complexities compared to disambiguating word senses when annotating text. These issues are discussed and illustrated, including the distinction between word senses and iconographic senses.

Our knowledge regarding the neural processing of the three chemical senses has been lagging behind that of our other senses considerably. It is only during the last 25 years that significant advances have been made in our understanding of where in the human brain odors, tastants, and trigeminal stimuli are processed. Here we provide an overview of the current knowledge of how the human brain pr...

Although it is commonly assumed that word sense disambiguation (WSD) should help to improve lexical choice and improve the quality of machine translation systems, how to successfully integrate word senses into such systems remains an unanswered question. Some successful approaches have involved reformulating either WSD or the word senses it produces, but work on using traditional word senses to...

This paper discusses certain problems arising within the treatment of the senses of functions in Alonzo Church’s Logic of Sense and Denotation. Church understands such senses themselves to be “sense-functions,” functions from sense to sense. However, the conditions he lays out under which a sense-function is to be regarded as a sense presenting another function as denotation allow for certain u...

We present a method for detecting word sense changes by utilizing automatically induced word senses. Our method works on the level of individual senses and allows a word to have e.g. one stable sense and then add a novel sense that later experiences change. Senses are grouped based on polysemy to find linguistic concepts and we can find broadening and narrowing as well as novel (polysemous and ...

In this paper, we propose a lexical senses representation system called E-HowNet, in which the lexical senses are defined by basic concepts. As a result, the meanings of expressions are more specific than those derived by using primitives. We also design an ontology to express the taxonomic relations between concepts and the attributes of concepts. To establish the taxonomic relations between w...

In this study we explore topic modeling for the automatic detection of new senses of known words. We apply methods developed in previous work for English (Lau et al., 2012, 2014) on a recent case of new word sense induction in French, namely the appearence of the new meaning of gesture for the word « quenelle ». Our experiments illustrate the potential of this approach at learning word senses, ...

The work described in this paper was originally motivated by the need to map verbs associated with FrameNet 1.2 frames to appropriate WordNet 2.0 senses. As the work evolved, it became apparent that the developed method was applicable for a number of other tasks, including assignment of WordNet senses to word lists used in attitude and opinion analysis, and collapsing WordNet senses into coarse...

We study the task of learning to rank images given a text query, a problem that is complicated by the issue of multiple senses. That is, the senses of interest are typically the visually distinct concepts that a user wishes to retrieve. In this paper, we propose to learn a ranking function that optimizes the ranking cost of interest and simultaneously discovers the disambiguated senses of the q...

This paper describes a heuristic approach to automatically identifying which senses of a machinereadable dictionary (MRD) headword are semantically related versus those which correspond to fundamentally different senses of the word. The inclusion of this information in a lexical database profoundly alters the nature of sense disambiguation: the appropriate "sense" of a polysemous word may now c...