Nowadays, software systems tend to include Artificial Intelligence (AI) components. Changes in the operational environment have been known to negatively impact the stability of AI-enabled software systems by causing unintended changes in behavior. However, how an environment configuration impacts the behavior of such systems has yet to be explored. Understanding and quantifying the degree of instability caused by different environment settings can help practitioners decide the best environment configuration for the most stable AI systems. To achieve this goal, we performed experiments with eight different combinations of three key environment variables (operating system, Python version, and CPU architecture) on 30 open-source AI-enabled systems using the Travis CI platform. We determine the existence and the degree of instability introduced by each configuration using three metrics: the output of an AI component of the system (model performance), the time required to build and run the system (processing time), and the cost associated with building and running the system (expense). Our results indicate that changes in environment configurations lead to instability across all three metrics; however, it is observed more frequently with respect to processing time and expense rather than model performance. For example, between Linux and MacOS, instability is observed in 23%, 96.67%, and 100% of the studied projects in model performance, processing time, and expense, respectively. Our findings underscore the importance of identifying the optimal combination of configuration settings to mitigate drops in model performance and reduce the processing time and expense before deploying an AI-enabled system.

Machine Learning (ML) systems are gaining popularity, reshaping various domains ranging from customer services to software engineering. The effectiveness of ML systems is dependent on the quality of their training data. Therefore, practitioners invest substantial time experimenting with different data, parameters, and models to guarantee the quality of the end system. Prior work highlighted unique challenges of developing ML systems, particularly concerning versioning data and models. Recently, various tools such as DVC and MLFlow have emerged to aid developers in the storage and tracking of data. Despite their growing popularity, very little is known about their usage patterns and impact on open-source software (OSS) systems. To address this gap, we conducted an empirical study on 56 GitHub OSS projects that use DVC to understand the DVC usage pattern and the impact of using DVC on the software development process. We found that Versioning and tracking is the most adopted DVC feature, being utilized by all 56 projects and being the only adopted feature in 85.7% of them. Furthermore, we found that DVC has a significant impact on the software development process indicators such as the number of created pull requests (PRs), and the number of bug-fix commits. For instance, our findings showed that DVC causes a peak in the number of commits and PRs at the moment of the adoption, followed by a long-term decrease. We believe that our findings can assist practitioners in tailoring tools to better meet user requirements and help organizations realize potential outcomes of adopting such tools.

Recent works have concluded that software code is more repetitive and predictable, i.e. more natural, than English texts. On re-examination, we find that much of the apparent "naturalness" of source code is due to the presence of language specific syntax, especially separators, such as semi-colons and brackets. For example, separators account for 44% of all tokens in our Java corpus. When we follow the NLP practices of eliminating punctuation (e.g., separators) and stopwords (e.g., keywords), we find that code is still repetitive and predictable, but to a lesser degree than previously thought. We suggest that SyntaxTokens be filtered to reduce noise in code recommenders. Unlike the code written for a particular project, API code usage is similar across projects: a file is opened and closed in the same manner regardless of domain. When we restrict our n-grams to those contained in the Java API, we find that API usages are highly repetitive. Since API calls are common across programs, researchers have made reliable statistical models to recommend sophisticated API call sequences. Sequential n-gram models were developed for natural languages. Code is usually represented by an AST which contains control and data flow, making n-grams models a poor representation of code. Comparing n-grams to statistical graph representations of the same codebase, we find that graphs are more repetitive and contain higherlevel patterns than n-grams. We suggest that future work focus on statistical code graphs models that accurately capture complex coding patterns. Our replication package makes our scripts and data available to future researchers[1].

Generating source code API sequences from an English query using Machine Translation (MT) has gained much interest in recent years. For any kind of MT, the model needs to be trained on a parallel corpus. In this paper we clean StackOverflow, one of the most popular online discussion forums for programmers, to generate a parallel English-Code corpus from Android posts. We contrast three data cleaning approaches: standard NLP, title only, and software task extraction. We evaluate the quality of the each corpus for MT. To provide indicators of how useful each corpus will be for machine translation, we provide researchers with measurements of the corpus size, percentage of unique tokens, and per-word maximum likelihood alignment entropy. We have used these corpus cleaning approaches to translate between English and Code [22, 23], to compare existing SMT approaches from word mapping to neural networks [24], and to re-examine the "natural software" hypothesis [29]. After cleaning and aligning the data, we create a simple maximum likelihood MT model to show that English words in the corpus map to a small number of specific code elements. This model provides a basis for the success of using StackOverflow for search and other tasks in the software engineering literature and paves the way for MT. Our scripts and corpora are publicly available on GitHub [1] as well as at https://search.datacite.org/works/10.5281/zenodo.2558551.