Stereotype-based semantic expansion for image retrieval

Abstract

We present a novel, stereotype-based semantic expansion approach to identify various image sets that stereotypically represent different aspects of a given keyword. Specifically, given an adjective keyword query, our method expands it to a set of noun sub-keywords, which are stereotypical examples that can be described by the given adjective (e.g., "cute" to "{infant, kitten, ...}"). We also perform a similar process for given noun keywords with adjectives (e.g., "infant" to "{cute, sweet, ...}"). To perform such expansion, we use Google Books n-grams, a new corpus of 500 million digitized books. We harvest stereotypical relationships among nouns and adjectives by utilizing useful lexical patterns such as similes on n-grams. In order to demonstrate benefits of our method, we have applied our method to image retrieval. By suggesting our expanded sub-keywords additionally to commonly co-occurring terms our method can explore unusual concepts and their corresponding images that are stereotypically related to the keyword.