Constructing Cooking Ontology for Live Streams

Abstract

We build a cooking domain knowledge by using an ontology schema that reflects natural language processing and enhances ontology instances with semantic query. Our research helps audiences to better understand live streaming, especially when they just switch to a show. The practical contribution of our research is to use cooking ontology, so we may map clips of cooking live stream video and instructions of recipes. The architecture of our study presents three sections: ontology construction, ontology enhancement, and mapping cooking video to cooking ontology. Also, our preliminary evaluations consist of three hierarchies-nodes, ordered-pairs, and 3-tuples-that we use to referee (1) ontology enhancement performance for our first experiment evaluation and (2) the accuracy ratio of mapping between video clips and cooking ontology for our second experiment evaluation. Our results indicate that ontology enhancement is effective and heightens accuracy ratios on matching pairs with cooking ontology and video clips.