Jianfeng Chen

Bytedance-Seed-Foundation-Code-Team, Yao Cheng, Jianfeng Chen et al. · bytedance

As the capabilities of code large language models (LLMs) continue to expand, their applications across diverse code intelligence domains are rapidly increasing. However, most existing datasets only evaluate limited application domains. To address this gap, we have developed a comprehensive code evaluation dataset FullStack Bench focusing on full-stack programming, which encompasses a wide range of application domains (e.g., basic programming, data analysis, software engineering, mathematics, and machine learning). Besides, to assess multilingual programming capabilities, in FullStack Bench, we design real-world instructions and corresponding unit test cases from 16 widely-used programming languages to reflect real-world usage scenarios rather than simple translations. Moreover, we also release an effective code sandbox execution tool (i.e., SandboxFusion) supporting various programming languages and packages to evaluate the performance of our FullStack Bench efficiently. Comprehensive experimental results on our FullStack Bench demonstrate the necessity and effectiveness of our FullStack Bench and SandboxFusion.

Jisheng Bai, Mou Wang, Haohe Liu et al.

Acoustic scene classification (ASC) is a crucial research problem in computational auditory scene analysis, and it aims to recognize the unique acoustic characteristics of an environment. One of the challenges of the ASC task is the domain shift between training and testing data. Since 2018, ASC challenges have focused on the generalization of ASC models across different recording devices. Although this task, in recent years, has achieved substantial progress in device generalization, the challenge of domain shift between different geographical regions, involving discrepancies such as time, space, culture, and language, remains insufficiently explored at present. In addition, considering the abundance of unlabeled acoustic scene data in the real world, it is important to study the possible ways to utilize these unlabelled data. Therefore, we introduce the task Semi-supervised Acoustic Scene Classification under Domain Shift in the ICME 2024 Grand Challenge. We encourage participants to innovate with semi-supervised learning techniques, aiming to develop more robust ASC models under domain shift.

Dongzhe Zhang, Jianfeng Chen, Jisheng Bai et al.

Deep learning-based sound event localization and classification is an emerging research area within wireless acoustic sensor networks. However, current methods for sound event localization and classification typically rely on a single microphone array, making them susceptible to signal attenuation and environmental noise, which limits their monitoring range. Moreover, methods using multiple microphone arrays often focus solely on source localization, neglecting the aspect of sound event classification. In this paper, we propose a deep learning-based method that employs multiple features and attention mechanisms to estimate the location and class of sound source. We introduce a Soundmap feature to capture spatial information across multiple frequency bands. We also use the Gammatone filter to generate acoustic features more suitable for outdoor environments. Furthermore, we integrate attention mechanisms to learn channel-wise relationships and temporal dependencies within the acoustic features. To evaluate our proposed method, we conduct experiments using simulated datasets with different levels of noise and size of monitoring areas, as well as different arrays and source positions. The experimental results demonstrate the superiority of our proposed method over state-of-the-art methods in both sound event classification and sound source localization tasks. And we provide further analysis to explain the reasons for the observed errors.

Wenqiang Sun, Han Yin, Jisheng Bai et al.

Suicide is one of the leading causes of death among adolescents. Previous suicide risk prediction studies have primarily focused on either textual or acoustic information in isolation, the integration of multimodal signals, such as speech and text, offers a more comprehensive understanding of an individual's mental state. Motivated by this, and in the context of the 1st SpeechWellness detection challenge, we explore a lightweight multi-branch multimodal system based on a dynamic fusion mechanism for speechwellness detection. To address the limitation of prior approaches that rely on time-domain waveforms for acoustic analysis, our system incorporates both time-domain and time-frequency (TF) domain acoustic features, as well as semantic representations. In addition, we introduce a dynamic fusion block to adaptively integrate information from different modalities. Specifically, it applies learnable weights to each modality during the fusion process, enabling the model to adjust the contribution of each modality. To enhance computational efficiency, we design a lightweight structure by simplifying the original baseline model. Experimental results demonstrate that the proposed system exhibits superior performance compared to the challenge baseline, achieving a 78% reduction in model parameters and a 5% improvement in accuracy.

Songqin Nong, Jingxuan Xu, Sheng Zhou et al.

As autonomous agents become adept at understanding and interacting with graphical user interface (GUI) environments, a new era of automated task execution is emerging. Recent studies have demonstrated that Reinforcement Learning (RL) can effectively enhance agents' performance in dynamic interactive GUI environments. However, these methods face two key limitations: (1) they overlook the significant variation in difficulty across different GUI tasks by treating the entire training data as a uniform set, which hampers the agent's ability to adapt its learning process; and (2) most approaches collapse task-specific nuances into a single, coarse reward, leaving the agent with a uniform signal that yields inefficient policy updates. To address these limitations, we propose CRAFT-GUI, a curriculum learning framework based on Group Relative Policy Optimization (GRPO) that explicitly accounts for the varying difficulty across trajectories. To enable more fine-grained policy optimization, we design a reward function that combines simple rule-based signals with model-judged evaluation, providing richer and more nuanced feedback during training. Experimental results demonstrate that our method achieves significant improvements over previous state-of-the-art approaches, outperforming them by 5.6% on public benchmarks Android Control and 10.3% on our internal online benchmarks, respectively. These findings empirically validate the effectiveness of integrating reinforcement learning with curriculum learning in GUI interaction tasks.

Jianfeng Chen, Xipeng Shen, Tim Menzies

Theorem provers has been used extensively in software engineering for software testing or verification. However, software is now so large and complex that additional architecture is needed to guide theorem provers as they try to generate test suites. The SNAP test suite generator (introduced in this paper) combines the Z3 theorem prover with the following tactic: cluster some candidate tests, then search for valid tests by proposing small mutations to the cluster centroids. This technique effectively removes repeated structures in the tests since many repeated structures can be replaced with one centroid. In practice, SNAP is remarkably effective. For 27 real-world programs with up to half a million variables, SNAP found test suites which were 10 to 750 smaller times than those found by the prior state-of-the-art. Also, SNAP ran orders of magnitude faster and (unlike prior work) generated 100% valid tests.

Xueqi Yang, Jianfeng Chen, Rahul Yedida et al.

Static code warning tools often generate warnings that programmers ignore. Such tools can be made more useful via data mining algorithms that select the "actionable" warnings; i.e. the warnings that are usually not ignored. In this paper, we look for actionable warnings within a sample of 5,675 actionable warnings seen in 31,058 static code warnings from FindBugs. We find that data mining algorithms can find actionable warnings with remarkable ease. Specifically, a range of data mining methods (deep learners, random forests, decision tree learners, and support vector machines) all achieved very good results (recalls and AUC (TRN, TPR) measures usually over 95% and false alarms usually under 5%). Given that all these learners succeeded so easily, it is appropriate to ask if there is something about this task that is inherently easy. We report that while our data sets have up to 58 raw features, those features can be approximated by less than two underlying dimensions. For such intrinsically simple data, many different kinds of learners can generate useful models with similar performance. Based on the above, we conclude that learning to recognize actionable static code warnings is easy, using a wide range of learning algorithms, since the underlying data is intrinsically simple. If we had to pick one particular learner for this task, we would suggest linear SVMs (since, at least in our sample, that learner ran relatively quickly and achieved the best median performance) and we would not recommend deep learning (since this data is intrinsically very simple).

Rui Shu, Tianpei Xia, Jianfeng Chen et al.

Background: In order that the general public is not vulnerable to hackers, security bug reports need to be handled by small groups of engineers before being widely discussed. But learning how to distinguish the security bug reports from other bug reports is challenging since they may occur rarely. Data mining methods that can find such scarce targets require extensive optimization effort. Goal: The goal of this research is to aid practitioners as they struggle to optimize methods that try to distinguish between rare security bug reports and other bug reports. Method: Our proposed method, called Swift, is a dual optimizer that optimizes both learner and pre-processor options. Since this is a large space of options, Swift uses a technique called epsilon-dominance that learns how to avoid operations that do not significantly improve performance. Result: When compared to recent state-of-the-art results (from FARSEC which is published in TSE'18), we find that the Swift's dual optimization of both pre-processor and learner is more useful than optimizing each of them individually. For example, in a study of security bug reports from the Chromium dataset, the median recalls of FARSEC and Swift were 15.7% and 77.4%, respectively. For another example, in experiments with data from the Ambari project, the median recalls improved from 21.5% to 85.7% (FARSEC to SWIFT). Conclusion: Overall, our approach can quickly optimize models that achieve better recalls than the prior state-of-the-art. These increases in recall are associated with moderate increases in false positive rates (from 8% to 24%, median). For future work, these results suggest that dual optimization is both practical and useful.

Jianfeng Chen, Joymallya Chakraborty, Philip Clark et al.

Maintaining web-services is a mission-critical task where any down-time means loss of revenue and reputation (of being a reliable service provider). In the current competitive web services market, such a loss of reputation causes extensive loss of future revenue. To address this issue, we developed SPIKE, a data mining tool which can predict upcoming service breakdowns, half an hour into the future. Such predictions let an organization alert and assemble the tiger team to address the problem (e.g. by reconfiguring cloud hardware in order to reduce the likelihood of that breakdown). SPIKE utilizes (a) regression tree learning (with CART); (b) synthetic minority over-sampling (to handle how rare spikes are in our data); (c) hyperparameter optimization (to learn best settings for our local data) and (d) a technique we called "topology sampling" where training vectors are built from extensive details of an individual node plus summary details on all their neighbors. In the experiments reported here, SPIKE predicted service spikes 30 minutes into future with recalls and precision of 75% and above. Also, SPIKE performed relatively better than other widely-used learning methods (neural nets, random forests, logistic regression).

Jianfeng Chen, Xipeng Shen, Tim Menzies

Software is now so large and complex that additional architecture is needed to guide theorem provers as they try to generate test suites. For example, the SNAP test suite generator (introduced in this paper) combines the Z3 theorem prover with the following tactic: sample around the average values seen in a few randomly selected valid tests. This tactic is remarkably effective. For 27 real-world programs with up to half a million variables, SNAP found test suites which were 10 to 750 smaller times than those found by the prior state-of-the-art. Also, SNAP ran orders of magnitude faster and (unlike prior work) generated 100% valid tests.

Tianpei Xia, Rahul Krishna, Jianfeng Chen et al.

Software analytics has been widely used in software engineering for many tasks such as generating effort estimates for software projects. One of the "black arts" of software analytics is tuning the parameters controlling a data mining algorithm. Such hyperparameter optimization has been widely studied in other software analytics domains (e.g. defect prediction and text mining) but, so far, has not been extensively explored for effort estimation. Accordingly, this paper seeks simple, automatic, effective and fast methods for finding good tunings for automatic software effort estimation. We introduce a hyperparameter optimization architecture called OIL (Optimized Inductive Learning). We test OIL on a wide range of hyperparameter optimizers using data from 945 software projects. After tuning, large improvements in effort estimation accuracy were observed (measured in terms of standardized accuracy). From those results, we recommend using regression trees (CART) tuned by different evolution combine with default analogy-based estimator. This particular combination of learner and optimizers often achieves in a few hours what other optimizers need days to weeks of CPU time to accomplish. An important part of this analysis is its reproducibility and refutability. All our scripts and data are on-line. It is hoped that this paper will prompt and enable much more research on better methods to tune software effort estimators.

Tianpei Xia, Jianfeng Chen, George Mathew et al.

Industrial practitioners now face a bewildering array of possible configurations for effort estimation. How to select the best one for a particular dataset? This paper introduces OIL (short for optimized learning), a novel configuration tool for effort estimation based on differential evolution. When tested on 945 software projects, OIL significantly improved effort estimations, after exploring just a few configurations (just a few dozen). Further OIL's results are far better than two methods in widespread use: estimation-via-analogy and a recent state-of-the-art baseline published at TOSEM'15 by Whigham et al. Given that the computational cost of this approach is so low, and the observed improvements are so large, we conclude that SBSE should be a standard component of software effort estimation.

Vivek Nair, Amritanshu Agrawal, Jianfeng Chen et al.

This paper introduces Data-Driven Search-based Software Engineering (DSE), which combines insights from Mining Software Repositories (MSR) and Search-based Software Engineering (SBSE). While MSR formulates software engineering problems as data mining problems, SBSE reformulates SE problems as optimization problems and use meta-heuristic algorithms to solve them. Both MSR and SBSE share the common goal of providing insights to improve software engineering. The algorithms used in these two areas also have intrinsic relationships. We, therefore, argue that combining these two fields is useful for situations (a) which require learning from a large data source or (b) when optimizers need to know the lay of the land to find better solutions, faster. This paper aims to answer the following three questions: (1) What are the various topics addressed by DSE? (2) What types of data are used by the researchers in this area? (3) What research approaches do researchers use? The paper briefly sets out to act as a practical guide to develop new DSE techniques and also to serve as a teaching resource. This paper also presents a resource (tiny.cc/data-se) for exploring DSE. The resource contains 89 artifacts which are related to DSE, divided into 13 groups such as requirements engineering, software product lines, software processes. All the materials in this repository have been used in recent software engineering papers; i.e., for all this material, there exist baseline results against which researchers can comparatively assess their new ideas.

Jianfeng Chen, Tim Menzies

Cloud computing provides engineers or scientists a place to run complex computing tasks. Finding a workflow's deployment configuration in a cloud environment is not easy. Traditional workflow scheduling algorithms were based on some heuristics, e.g. reliability greedy, cost greedy, cost-time balancing, etc., or more recently, the meta-heuristic methods, such as genetic algorithms. These methods are very slow and not suitable for rescheduling in the dynamic cloud environment. This paper introduces RIOT (Randomized Instance Order Types), a stochastic based method for workflow scheduling. RIOT groups the tasks in the workflow into virtual machines via a probability model and then uses an effective surrogate-based method to assess a large amount of potential scheduling. Experiments in dozens of study cases showed that RIOT executes tens of times faster than traditional methods while generating comparable results to other methods.

Jianfeng Chen, Vivek Nair, Tim Menzies

Context: Evolutionary algorithms typically require a large number of evaluations (of solutions) to converge - which can be very slow and expensive to evaluate.Objective: To solve search-based software engineering (SE) problems, using fewer evaluations than evolutionary methods.Method: Instead of mutating a small population, we build a very large initial population which is then culled using a recursive bi-clustering chop approach. We evaluate this approach on multiple SE models, unconstrained as well as constrained, and compare its performance with standard evolutionary algorithms. Results: Using just a few evaluations (under 100), we can obtain comparable results to state-of-the-art evolutionary algorithms.Conclusion: Just because something works, and is widespread use, does not necessarily mean that there is no value in seeking methods to improve that method. Before undertaking search-based SE optimization tasks using traditional EAs, it is recommended to try other techniques, like those explored here, to obtain the same results with fewer evaluations.

Jianfeng Chen, Vivek Nair, Rahul Krishna et al.

Increasingly, Software Engineering (SE) researchers use search-based optimization techniques to solve SE problems with multiple conflicting objectives. These techniques often apply CPU-intensive evolutionary algorithms to explore generations of mutations to a population of candidate solutions. An alternative approach, proposed in this paper, is to start with a very large population and sample down to just the better solutions. We call this method "SWAY", short for "the sampling way". Sway is very simple to implement and, in studies with various software engineering models, this sampling approach was found to be competitive with corresponding state-of-the-art evolutionary algorithms while requiring far less computation cost. Considering the simplicity and effectiveness of Sway, we, therefore, propose this approach as a baseline method for search-based software engineering models, especially for models that are very slow to execute.