This article presents a study on semantic parsing of cooking recipes to make them understandable for IoT kitchen devices. The authors developed a unique annotated corpus aimed at transforming recipe instructions into commands for these devices. The study used machine learning-based sequence labelling methods to parse recipe instructions and extract tuples of interest from them. The results showed that while it is possible to train semantic parsers based on their annotations, most natural-language instructions are incomplete, making the transformation process challenging.
Publication date: 26 Sep 2023
Project Page: Not Provided
Paper: https://arxiv.org/pdf/2309.14165
