In the last decade, information and communication technologies have been highly developed. For convenience, many applications have been installed in smartphones instead of desktop computers. As a popular platform, Google Play presents thousands of mobile applications. Because there are so many dazzling applications, it is difficult for users to determine which are suitable for their needs. Many factors are likely to influence the purchase of an application, such as advertisements, word of mouth, and other media. In deciding whether to purchase an application, users probably refer to customer reviews. Indeed, users may take a significant amount of time to evaluate the legitimacy of the reviews. In this paper, we introduce a concept for recommending applications. Based on pointwise mutual information, we calculate the positive or negative score of semantic orientation in each review. We also consider subjective factors (i.e., public opinion, anonymous opinion, and star rating) and objective factors (i.e., download number and reputation).