Online information sources such as websites, search engines, digital maps, etc. provide restaurant information which often includes an affordability indicator that indicates the general level of prices charged by the restaurant for the items on its menu. Currently, such affordability indicators are derived from subjective and expensive consumer surveys, which can be prone to error.
This disclosure describes techniques to automatically infer an affordability indicator for a restaurant using data obtained from existing data sources and provided as input to a suitably trained machine learning model. Prices are obtained from sources of high-quality structured information and/or lower-quality unstructured information, such as user-generated content. A trained model utilizes the obtained price information and generates the affordability indicator. The model is designed to take into account local differences in pricing and cost of living via appropriate normalization that enables the operation to scale globally in a seamless manner.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Yang, Xinru, "Inferring Restaurant Affordability From Menu Prices", Technical Disclosure Commons, (July 29, 2021)