PAWSM food similarity tool searches for food with similar attributes. For this type of search, we use several food attributes. The main ones being types of food (dry, wet, raw and treat) and their caloric value.
For data clustering, we use the k-means method to get our food neatly organized in clusters with similar attributes.
The search and its results can be used when you're buying your dog's food and you cannot find your usual food. The output results are randomized.