Biblio

The current evaluation of API recommendation systems mainly focuses on correctness, which is calculated through matching results with ground-truth APIs. However, this measurement may be affected if there exist more than one APIs in a result. In practice, some APIs are used to implement basic functionalities (e.g., print and log generation). These APIs can be invoked everywhere, and they may contribute less than functionally related APIs to the given requirements in recommendation. To study the impacts of correct-but-useless APIs, we use utility to measure them. Our study is conducted on more than 5,000 matched results generated by two specification-based API recommendation techniques. The results show that the matched APIs are heavily overlapped, 10% APIs compose more than 80% matched results. The selected 10% APIs are all correct, but few of them are used to implement the required functionality. We further propose a heuristic approach to measure the utility and conduct an online evaluation with 15 developers. Their reports confirm that the matched results with higher utility score usually have more efforts on programming than the lower ones.

It is time-consuming and labor-intensive to learn and locate the correct API for programming tasks. Thus, it is beneficial to perform API recommendation automatically. The graph-based statistical model has been shown to recommend top-10 API candidates effectively. It falls short, however, in accurately recommending an actual top-1 API. To address this weakness, we propose RecRank, an approach and tool that applies a novel ranking-based discriminative approach leveraging API usage path features to improve top-1 API recommendation. Empirical evaluation on a large corpus of (1385+8) open source projects shows that RecRank significantly improves top-1 API recommendation accuracy and mean reciprocal rank when compared to state-of-the-art API recommendation approaches.

Learning and remembering how to use APIs is difficult. While code-completion tools can recommend API methods, browsing a long list of API method names and their documentation is tedious. Moreover, users can easily be overwhelmed with too much information. We present a novel API recommendation approach that taps into the predictive power of repetitive code changes to provide relevant API recommendations for developers. Our approach and tool, APIREC, is based on statistical learning from fine-grained code changes and from the context in which those changes were made. Our empirical evaluation shows that APIREC correctly recommends an API call in the first position 59% of the time, and it recommends the correct API call in the top five positions 77% of the time. This is a significant improvement over the state-of-the-art approaches by 30-160% for top-1 accuracy, and 10-30% for top-5 accuracy, respectively. Our result shows that APIREC performs well even with a one-time, minimal training dataset of 50 publicly available projects.