Composing and Embedding the Words-as-Classifiers Model of Grounded Semantics
This work addresses incremental improvements in semantic modeling for natural language processing researchers, focusing on combining grounded and distributional approaches.
The paper explored how the words-as-classifiers model of grounded semantics can perform composition and be unified with distributional representations, using classifier coefficients as embeddings and developing composition methods for different classifier types. Results on the refCOCO dataset showed the need to expand composition strategies and integrate grounded and distributional representations.
The words-as-classifiers model of grounded lexical semantics learns a semantic fitness score between physical entities and the words that are used to denote those entities. In this paper, we explore how such a model can incrementally perform composition and how the model can be unified with a distributional representation. For the latter, we leverage the classifier coefficients as an embedding. For composition, we leverage the underlying mechanics of three different classifier types (i.e., logistic regression, decision trees, and multi-layer perceptrons) to arrive at a several systematic approaches to composition unique to each classifier including both denotational and connotational methods of composition. We compare these approaches to each other and to prior work in a visual reference resolution task using the refCOCO dataset. Our results demonstrate the need to expand upon existing composition strategies and bring together grounded and distributional representations.