Recommendation systems have become increasingly popular in recent years. In today’s world where the competition is extremely high, many service providers, from e-commerce to music industry, use recommendation systems. Considering the number of products and customers, it is crucial to deliver the right products to the right customers. Numerous methods are applied in recommendation systems. In our study, it is aimed to examine the purchasing association between products and to offer alternative products to the customers based on what they have already bought. For this purpose, Apriori, and Word2Vec and FastText, a natural language processing method, which extract the relationship space between words according to their frequency of occurrence, are used. Then, these models are compared. The results from word embedding methods are promising and will be used in future studies.