As many new traders seek to earn their share in the rapidly emerging cryptocurrency domain, greater reliance is placed on digital currency exchanges to facilitate this significant demand. With malicious users establishing fake exchanges to commit fraudulent crimes, there is a great need to classify the trustworthiness of exchanges. Both research studies and practical applications have aimed to characterize features of credible exchanges, but may not be sufficient to reflect the perception of their trustworthiness. In this thesis, we introduce a metric for evaluating exchanges based on direct user sentiment. We explore the effectiveness of our metric by utilizing machine learning tactics to compare existing ranking lists to observe how well our ranking performs.