Systems and methods for recommending an item from a plurality of items. The plurality of items includes chemistry-based features. The system comprises a computing device and a database. The database includes user preference data, item data, and mapping data. The system receives user input regarding user preferences, and associates user preference values with chemistry-based feature values. The system compares chemistry-based feature values for each item to the user preference values and returns items that best match. In one embodiment, the items are wines and the system uses chemistry-based features of wines such as wine volatile chemical compounds to return wine recommendations.