This paper studies programmer visual attention on code as it relates to underlying function call graphs during code summarization. Programmer visual attention refers to where people look when performing a software engineering task, and code summarization is the task of writing a natural language description about a section of source code. Prior work has studied programmers’ visual attention during code summarization, with the vast majority of research effort placed on details in single functional units of code. There have not been any techniques developed to understand code comprehension at the project level due to the difficulty of this task, despite the nature of most real-world methods as embedded within complex project context. This paper focuses on the visual attention paid to the call graph context in which a method sits. We analyze visual attention coverage of call graphs with graph-based metrics, such as the depth that programmers traverse or the amount of coverage they attain. We use these metrics, among other means, to reevaluate an existing dataset from a previous eye-tracking study of programmers (n = 10) that considered basic properties of programmer visual attention in a project context. We then created a new dataset (n = 12) using the same procedures specifically for this paper, resulting in a total of 88 hours of recorded visual behavior on source code. We used our proposed metrics to analyze how participants’ visual strategies correlated with their code summary quality, and confidence in their summaries. Interestingly, we found that higher coverage of the call graph was associated with decreases in both summary quality and participants’ confidence.

Programmer attention represents the visual focus of programmers on parts of the source code in pursuit of programming tasks. The focus of current research in modeling this programmer attention has been on using mouse cursors, keystrokes, or eye tracking equipment to map areas in a snippet of code. These approaches have traditionally only mapped attention for a single method. However, there is a knowledge gap in the literature because programming tasks such as source code summarization require programmers to use contextual knowledge that can only be found in other parts of the project, not only in a single method. To address this knowledge gap, we conducted an in-depth human study with 10 Java programmers, where each programmer generated summaries for 40 methods from five large Java projects over five one-hour sessions. We used eye tracking equipment to map the visual attention of programmers while they wrote the summaries. We also rate the quality of each summary. We found eye-gaze patterns and metrics that define common behaviors between programmer attention during context-aware code summarization. Specifically, we found that programmers need to read up to 35% fewer words (p$\boldsymbol{ \lt }$0.01) over the whole session, and revisit 13% fewer words (p$ \lt $0.03) as they summarize each method during a session, while maintaining the quality of summaries. We also found that the amount of source code a participant looks at correlates with a higher quality summary, but this trend follows a bell-shaped curve, such that after a threshold reading more source code leads to a significant decrease (p$\boldsymbol{ \lt }$0.01) in the quality of summaries. We also gathered insight into the type of methods in the project that provide the most contextual information for code summarization based on programmer attention. Specifically, we observed that programmers spent a majority of their time looking at methods inside the same class as the target method to be summarized. Surprisingly, we found that programmers spent significantly less time looking at methods in the call graph of the target method. We discuss how our empirical observations may aid future studies towards modeling programmer attention and improving context-aware automatic source code summarization.

The increasing use of large language model (LLM)-powered code generation tools, such as GitHub Copilot, is transforming software engineering practices. This paper investigates how developers validate and repair code generated by Copilot and examines the impact of code provenance awareness during these processes. We conducted a lab study with 28 participants tasked with validating and repairing Copilot-generated code in three software projects. Participants were randomly divided into two groups: one informed about the provenance of LLM-generated code and the other not. We collected data on IDE interactions, eye-tracking, cognitive workload assessments, and conducted semi-structured interviews. Our results indicate that, without explicit information, developers often fail to identify the LLM origin of the code. Developers exhibit LLM-specific behaviors such as frequent switching between code and comments, different attentional focus, and a tendency to delete and rewrite code. Being aware of the code’s provenance led to improved performance, increased search efforts, more frequent Copilot usage, and higher cognitive workload. These findings enhance our understanding of developer interactions with LLM-generated code and inform the design of tools for effective human-LLM collaboration in software development.

Code summarization is the task of creating short, natural language descriptions of source code. It is an important part of code comprehension and a powerful method of documentation. Previous work has made progress in identifying where programmers focus in code as they write their own summaries (i.e., Writing). However, there is currently a gap in studying programmers’ attention as they read code with pre-written summaries (i.e., Reading). As a result, it is currently unknown how these two forms of code comprehension compare: Reading and Writing. Also, there is a limited understanding of programmer attention with respect to program semantics. We address these shortcomings with a human eye-tracking study ( n = 27) comparing Reading and Writing. We examined programmers’ attention with respect to fine-grained program semantics, including their attention sequences (i.e., scan paths). We find distinctions in programmer attention across the comprehension tasks, similarities in reading patterns between them, and differences mediated by demographic factors. This can help guide code comprehension in both computer science education and automated code summarization. Furthermore, we mapped programmers’ gaze data onto the Abstract Syntax Tree to explore another representation of human attention. We find that visual behavior on this structure is not always consistent with that on source code.

Label smoothing is a regularization technique for neural networks. Normally neural models are trained to an output distribution that is a vector with a single 1 for the correct prediction, and 0 for all other elements. Label smoothing converts the correct prediction location to something slightly less than 1, then distributes the remainder to the other elements such that they are slightly greater than 0. A conceptual explanation behind label smoothing is that it helps prevent a neural model from becoming "overconfident" by forcing it to consider alternatives, even if only slightly. Label smoothing has been shown to help several areas of language generation, yet typically requires considerable tuning and testing to achieve the optimal results. This tuning and testing has not been reported for neural source code summarization – a growing research area in software engineering that seeks to generate natural language descriptions of source code behavior. In this paper, we demonstrate the effect of label smoothing on several baselines in neural code summarization, and conduct an experiment to find good parameters for label smoothing and make recommendations for its use.

This paper launches a new effort at modeling programmer attention by predicting eye movement scanpaths. Programmer attention refers to what information people intake when performing programming tasks. Models of programmer attention refer to machine prediction of what information is important to people. Models of programmer attention are important because they help researchers build better interfaces, assistive technologies, and more human-like AI. For many years, researchers in SE have built these models based on features such as mouse clicks, key logging, and IDE interactions. Yet the holy grail in this area is scanpath prediction - the prediction of the sequence of eye fixations a person would take over a visual stimulus. A person's eye movements are considered the most concrete evidence that a person is taking in a piece of information. Scanpath prediction is a notoriously difficult problem, but we believe that the emergence of lower-cost, higheraccuracy eye tracking equipment and better large language models of source code brings a solution within grasp. We present an eye tracking experiment with 27 programmers and a prototype scanpath predictor to present preliminary results and obtain early community feedback.

This tool demonstration presents a research toolkit for a language model of Java source code. The target audience includes researchers studying problems at the granularity level of subroutines, statements, or variables in Java. In contrast to many existing language models, we prioritize features for researchers including an open and easily-searchable training set, a held out test set with different levels of deduplication from the training set, infrastructure for deduplicating new examples, and an implementation platform suitable for execution on equipment accessible to a relatively modest budget. Our model is a GPT2-like architecture with 350m parameters. Our training set includes 52m Java methods (9b tokens) and 13m StackOverflow threads (10.5b tokens). To improve accessibility of research to more members of the community, we limit local resource requirements to GPUs with 16GB video memory. We provide a test set of held out Java methods that include descriptive comments, including the entire Java projects for those methods. We also provide deduplication tools using precomputed hash tables at various similarity thresholds to help researchers ensure that their own test examples are not in the training set. We make all our tools and data open source and available via Huggingface and Github.

Neural source code summarization is the task of generating natural language descriptions of source code behavior using neural networks. A fundamental component of most neural models is an attention mechanism. The attention mechanism learns to connect features in source code to specific words to use when generating natural language descriptions. Humans also pay attention to some features in code more than others. This human attention reflects experience and high-level cognition well beyond the capability of any current neural model. In this paper, we use data from published eye-tracking experiments to create a model of this human attention. The model predicts which words in source code are the most important for code summarization. Next, we augment a baseline neural code summarization approach using our model of human attention. We observe an improvement in prediction performance of the augmented approach in line with other bio-inspired neural models.

Source code summarization is the task of writing natural language descriptions of source code. The primary use of these descriptions is in documentation for programmers. Automatic generation of these descriptions is a high value research target due to the time cost to programmers of writing these descriptions themselves. In recent years, a confluence of software engineering and artificial intelligence research has made inroads into automatic source code summarization through applications of neural models of that source code. However, an Achilles’ heel to a vast majority of approaches is that they tend to rely solely on the context provided by the source code being summarized. But empirical studies in program comprehension are quite clear that the information needed to describe code much more often resides in the context in the form of Function Call Graph surrounding that code. In this paper, we present a technique for encoding this call graph context for neural models of code summarization. We implement our approach as a supplement to existing approaches, and show statistically significant improvement over existing approaches. In a human study with 20 programmers, we show that programmers perceive generated summaries to generally be as accurate, readable, and concise as human-written summaries.

Source code summarization involves creating brief descriptions of source code in natural language. These descriptions are a key component of software documentation such as JavaDocs. Automatic code summarization is a prized target of software engineering research, due to the high value summaries have to programmers and the simultaneously high cost of writing and maintaining documentation by hand. Current work is almost all based on machine models trained via big data input. Large datasets of examples of code and summaries of that code are used to train an e.g. encoder-decoder neural model. Then the output predictions of the model are evaluated against a set of reference summaries. The input is code not seen by the model, and the prediction is compared to a reference. The means by which a prediction is compared to a reference is essentially word overlap, calculated via a metric such as BLEU or ROUGE. The problem with using word overlap is that not all words in a sentence have the same importance, and many words have synonyms. The result is that calculated similarity may not match the perceived similarity by human readers. In this paper, we conduct an experiment to measure the degree to which various word overlap metrics correlate to human-rated similarity of predicted and reference summaries. We evaluate alternatives based on current work in semantic similarity metrics and propose recommendations for evaluation of source code summarization.

Virtual Assistant technology is rapidly proliferating to improve productivity in a variety of tasks. While several virtual assistants for everyday tasks are well-known (e.g., Siri, Cortana, Alexa), assistants for specialty tasks such as software engineering are rarer. One key reason software engineering assistants are rare is that very few experimental datasets are available and suitable for training the AI that is the bedrock of current virtual assistants. In this paper, we present a set of Wizard of Oz experiments that we designed to build a dataset for creating a virtual assistant. Our target is a hypothetical virtual assistant for helping programmers use APIs. In our experiments, we recruited 30 professional programmers to complete programming tasks using two APIs. The programmers interacted with a simulated virtual assistant for help – the programmers were not aware that the assistant was actually operated by human experts. We then annotated the dialogue acts in the corpus along four dimensions: illocutionary intent, API information type(s), backward-facing function, and traceability to specific API components. We observed a diverse range of interactions that will facilitate the development of dialogue strategies for virtual assistants for API usage.

A source code summary of a subroutine is a brief description of that subroutine. Summaries underpin a majority of documentation consumed by programmers, such as the method summaries in JavaDocs. Source code summarization is the task of writing these summaries. At present, most state-of-the-art approaches for code summarization are neural network-based solutions akin to seq2seq, graph2seq, and other encoder-decoder architectures. The input to the encoder is source code, while the decoder helps predict the natural language summary. While these models tend to be similar in structure, evidence is emerging that different models make different contributions to prediction quality - differences in model performance are orthogonal and complementary rather than uniform over the entire dataset. In this paper, we explore the orthogonal nature of different neural code summarization approaches and propose ensemble models to exploit this orthogonality for better overall performance. We demonstrate that a simple ensemble strategy boosts performance by up to 14.8%, and provide an explanation for this boost. The takeaway from this work is that a relatively small change to the inference procedure in most neural code summarization techniques leads to outsized improvements in prediction quality.

Source code summarization of a subroutine is the task of writing a short, natural language description of that subroutine. The description usually serves in documentation aimed at programmers, where even brief phrase (e.g. "compresses data to a zip file") can help readers rapidly comprehend what a subroutine does without resorting to reading the code itself. Techniques based on neural networks (and encoder-decoder model designs in particular) have established themselves as the state-of-the-art. Yet a problem widely recognized with these models is that they assume the information needed to create a summary is present within the code being summarized itself - an assumption which is at odds with program comprehension literature. Thus a current research frontier lies in the question of encoding source code context into neural models of summarization. In this paper, we present a project-level encoder to improve models of code summarization. By project-level, we mean that we create a vectorized representation of selected code files in a software project, and use that representation to augment the encoder of state-of-the-art neural code summarization techniques. We demonstrate how our encoder improves several existing models, and provide guidelines for maximizing improvement while controlling time and resource costs in model size.

A question answering (QA) system is a type of conversational AI that generates natural language answers to questions posed by human users. QA systems often form the backbone of interactive dialogue systems, and have been studied extensively for a wide variety of tasks ranging from restaurant recommendations to medical diagnostics. Dramatic progress has been made in recent years, especially from the use of encoder-decoder neural architectures trained with big data input. In this paper, we take initial steps to bringing state-of-the-art neural QA technologies to Software Engineering applications by designing a context-based QA system for basic questions about subroutines. We curate a training dataset of 10.9 million question/context/answer tuples based on rules we extract from recent empirical studies. Then, we train a custom neural QA model with this dataset and evaluate the model in a study with professional programmers. We demonstrate the strengths and weaknesses of the system, and lay the groundwork for its use in eventual dialogue systems for software engineering.

Source code summarization is the task of creating short, natural language descriptions of source code. Code summarization is the backbone of much software documentation such as JavaDocs, in which very brief comments such as "adds the customer object" help programmers quickly understand a snippet of code. In recent years, automatic code summarization has become a high value target of research, with approaches based on neural networks making rapid progress. However, as we will show in this paper, the production of good summaries relies on the production of the action word in those summaries: the meaning of the example above would be completely changed if "removes" were substituted for "adds." In this paper, we advocate for a special emphasis on action word prediction as an important stepping stone problem towards better code summarization - current techniques try to predict the action word along with the whole summary, and yet action word prediction on its own is quite difficult. We show the value of the problem for code summaries, explore the performance of current baselines, and provide recommendations for future research.