Search

Developers often perform repetitive code editing activities (up to 70%) for various reasons (e.g., code refactoring) during software development. Many deep learning (DL) models have been proposed to automate code editing by learning from the code editing history. Among DL-based models, pre-trained code editing models have achieved the state-of-the-art (SOTA) results. Pre-trained models are first pre-trained with pre-training tasks and fine-tuned with the code editing task. Existing pre-training tasks mainly are code infilling tasks (e.g., masked language modeling), which are derived from the natural language processing field and are not designed for automatic code editing. In this article, we propose a novel pre-training task specialized in code editing and present an effective pretrained code editing model named CodeEditor. Compared to previous code infilling tasks, our pre-training task further improves the performance and generalization ability of code editing models. Specifically, we collect lots of real-world code snippets as the ground truth and use a powerful generator to rewrite them into mutated versions. Then, we pre-train our CodeEditor to edit mutated versions into the corresponding ground truth, to learn edit patterns. We conduct experiments on four code editing datasets and evaluate the pre-trained CodeEditor in three settings (i.e., fine-tuning, few-shot, and zero-shot). (1) In the fine-tuning setting, we train the pre-trained CodeEditor with four datasets and evaluate it on the test data. CodeEditor outperforms the SOTA baselines by 15%, 25.5%, 9.4%, and 26.6% on four datasets. (2) In the few-shot setting, we train the pre-trained CodeEditor with limited data and evaluate it on the test data. CodeEditor substantially performs better than all baselines, even outperforming baselines that are fine-tuned with all data. (3) In the zero-shot setting, we evaluate the pre-trained CodeEditor on the test data without training. CodeEditor correctly edits 1,113 programs, while the SOTA baselines cannot work. The results show that the superiority of our pre-training task and the pre-trained CodeEditor is more effective in automatic code editing. [ABSTRACT FROM AUTHOR]

Deep Learning (DL) techniques for Natural Language Processing have been evolving remarkably fast. Recently, the DL advances in language modeling, machine translation, and paragraph understanding are so prominent that the potential of DL in Software Engineering cannot be overlooked, especially in the field of program learning. To facilitate further research and applications of DL in this field, we provide a comprehensive review to categorize and investigate existing DL methods for source code modeling and generation. To address the limitations of the traditional source code models, we formulate common program learning tasks under an encoder-decoder framework. After that, we introduce recent DL mechanisms suitable to solve such problems. Then, we present the state-of-the-art practices and discuss their challenges with some recommendations for practitioners and researchers as well. [ABSTRACT FROM AUTHOR]

The next significant step in the evolution and proliferation of artificial intelligence technology will be the integration of neural network (NN) models within embedded and mobile systems. This calls for the design of compact, energy efficient NN models in silicon. In this article, we present a scalable applicationspecific integrated circuit (ASIC) design of an energy-efficient Long Short-Term Memory (LSTM) accelerator, named ELSA, which is suitable for energy-constrained devices. It includes several architectural innovations to achieve small area and high energy efficiency. To reduce the area and power consumption of the overall design, the compute-intensive units of ELSA employ approximate multiplications and still achieve high performance and accuracy. The performance is further improved through efficient synchronization of the elastic pipeline stages to maximize the utilization. The article also includes a performance model of ELSA, as a function of the hidden nodes and timesteps, permitting its use for the evaluation of any LSTM application. ELSA was implemented in register transfer level (RTL) and was synthesized and placed and routed in 65nm technology. Its functionality is demonstrated for language modeling--a common application of LSTM. ELSA is compared against a baseline implementation of an LSTM accelerator with standard functional units and without any of the architectural innovations of ELSA. The article demonstrates that ELSA can achieve significant improvements in power, area, and energy-efficiency when compared to the baseline design and several ASIC implementations reported in the literature, making it suitable for use in embedded systems and real-time applications. [ABSTRACT FROM AUTHOR]

Ever-growing electronic medical corpora provide unprecedented opportunities for researchers to analyze patient conditions and drug effects. Meanwhile, severe challenges emerged in the large-scale electronic medical records process phase. Primarily, emerging words for medical terms, including informal descriptions, are difficult to recognize. Moreover, although deep models can help in entity extraction on medical texts, they require large-scale labels, which are time-intensive to obtain and not always available in the medical domain. However, when encountering a situation where massive unseen concepts appear or labeled data is insufficient, the performance of existing algorithms will suffer an intolerable decline. In this article, we propose a balanced and deep active learning framework for Medical Named Entity Recognition (MedNER) to alleviate the above problems. Specifically, to describe our selection strategy precisely, we first define the uncertainty of a medical sentence as a labeling loss predicted by a loss-prediction module and define diversity as the least text distance between pairs of sentences in a sample batch computed based on word-morpheme embeddings. Furthermore, aiming to make a trade-off between uncertainty and diversity, we formulate a Distinct-K optimization problem to maximize the slightest uncertainty and diversity of chosen sentences. Finally, we propose a threshold-based approximation selection algorithm, Distinct-K Filter, which selects the most beneficial training samples by balancing diversity and uncertainty. Extensive experimental results on real datasets demonstrate that MedNER significantly outperforms existing approaches. [ABSTRACT FROM AUTHOR]

Code models have made significant advancements in code intelligence by encoding knowledge about programming languages. While previous studies have explored the capabilities of these models in learning code syntax, there has been limited investigation on their ability to understand code semantics. Additionally, existing analyses assume that the number of edges between nodes at the abstract syntax tree (AST) is related to syntax distance, and also often require transforming the high-dimensional space of deep learning models to a low-dimensional one, which may introduce inaccuracies. To study how code models represent code syntax and semantics, we conduct a comprehensive analysis of seven code models, including four representative code pre-trained models (CodeBERT, GraphCodeBERT, CodeT5, and UnixCoder) and three large language models (LLMs) (StarCoder, CodeLlama and CodeT5+). We design four probing tasks to assess the models' capacities in learning both code syntax and semantics. These probing tasks reconstruct code syntax and semantics structures (AST, control dependence graph (CDG), data dependence graph (DDG), and control flow graph (CFG)) in the representation space. These structures are core concepts for code understanding. We also investigate the syntax token role in each token representation and the long dependency between the code tokens. Additionally, we analyze the distribution of attention weights related to code semantic structures. Through extensive analysis, our findings highlight the strengths and limitations of different code models in learning code syntax and semantics. The results demonstrate that these models excel in learning code syntax, successfully capturing the syntax relationships between tokens and the syntax roles of individual tokens. However, their performance in encoding code semantics varies. CodeT5 and CodeBERT demonstrate proficiency in capturing control and data dependencies, whereas UnixCoder shows weaker performance in this aspect. We do not observe LLMs generally performing much better than pre-trained models. The shallow layers of LLMs perform better than their deep layers. The investigation of attention weights reveals that different attention heads play distinct roles in encoding code semantics. Our research findings emphasize the need for further enhancements in code models to better learn code semantics. This study contributes to the understanding of code models' abilities in syntax and semantics analysis. Our findings provide guidance for future improvements in code models, facilitating their effective application in various code-related tasks. [ABSTRACT FROM AUTHOR]

(Source) code search is widely concerned by software engineering researchers because it can improve the productivity and quality of software development. Given a functionality requirement usually described in a natural language sentence, a code search system can retrieve code snippets that satisfy the requirement from a large-scale code corpus, e.g., GitHub. To realize effective and efficient code search, many techniques have been proposed successively. These techniques improve code search performance mainly by optimizing three core components, including query understanding component, code understanding component, and query-code matching component. In this article, we provide a 3-dimensional perspective survey for code search. Specifically, we categorize existing code search studies into query-end optimization techniques, code-end optimization techniques, and match-end optimization techniques according to the specific components they optimize. These optimization techniques are proposed to enhance the performance of specific components, and thus the overall performance of code search. Considering that each end can be optimized independently and contributes to the code search performance, we treat each end as a dimension. Therefore, this survey is 3-dimensional in nature, and it provides a comprehensive summary of each dimension in detail. To understand the research trends of the three dimensions in existing code search studies, we systematically review 68 relevant literatures. Different from existing code search surveys that only focus on the query end or code end or introduce various aspects shallowly (including codebase, evaluation metrics, modeling technique, etc.), our survey provides a more nuanced analysis and review of the evolution and development of the underlying techniques used in the three ends. Based on a systematic review and summary of existing work, we outline several open challenges and opportunities at the three ends that remain to be addressed in future work. [ABSTRACT FROM AUTHOR]

With the rapid development of deep learning (DL), the recent trend of log-based anomaly detection focuses on extracting semantic information from log events (i.e., templates of log messages) and designing more advanced DL models for anomaly detection. Indeed, the effectiveness of log-based anomaly detection can be improved, but these DL-based techniques further suffer from the limitations of more heavy dependency on training data (such as data quality or data labels) and higher costs in time and resources due to the complexity and scale of DL models, which hinder their practical use. On the contrary, the techniques based on traditional machine learning or data mining algorithms are less dependent on training data and more efficient but produce worse effectiveness than DL-based techniques, which is mainly caused by the problem of unseen log events (some log events in incoming log messages are unseen in training data) confirmed by our motivating study. Intuitively, if we can improve the effectiveness of traditional techniques to be comparable with advanced DL-based techniques, then log-based anomaly detection can be more practical. Indeed, an existing study in the other area (i.e., linking questions posted on Stack Overflow) has pointed out that traditional techniques with some optimizations can indeed achieve comparable effectiveness with the state-of-the-art DL-based technique, indicating the feasibility of enhancing traditional log-based anomaly detection techniques to some degree. Inspired by the idea of "try-with-simpler," we conducted the first empirical study to explore the potential of improving traditional techniques for more practical log-based anomaly detection. In this work, we optimized the traditional unsupervised PCA (Principal Component Analysis) technique by incorporating a lightweight semantic-based log representation in it, called SemPCA, and conducted an extensive study to investigate the potential of SemPCA for more practical log-based anomaly detection. By comparing seven log-based anomaly detection techniques (including four DL-based techniques, two traditional techniques, and SemPCA) on both public and industrial datasets, our results show that SemPCA achieves comparable effectiveness as advanced supervised/semi-supervised DL-based techniques while being much more stable under insufficient training data and more efficient, demonstrating that the traditional technique can still excel after small but useful adaptation. [ABSTRACT FROM AUTHOR]

A networked time series (NETS) is a family of time series on a given graph, one for each node. It has a wide range of applications from intelligent transportation to environment monitoring to smart grid management. An important task in such applications is to predict the future values of a NETS based on its historical values and the underlying graph. Most existing methods require complete data for training. However, in real-world scenarios, it is not uncommon to have missing data due to sensor malfunction, incomplete sensing coverage, and so on. In this article, we study the problem of NETS prediction with incomplete data. We propose networked time series Imputation Generative Adversarial Network (NETS-ImpGAN), a novel deep learning framework that can be trained on incomplete data with missing values in both history and future. Furthermore, we propose Graph Temporal Attention Networks, which incorporate the attention mechanism to capture both inter-time series and temporal correlations. We conduct extensive experiments on four real-world datasets under different missing patterns and missing rates. The experimental results show that NETS-ImpGAN outperforms existing methods, reducing the Mean Absolute Error by up to 25%. [ABSTRACT FROM AUTHOR]

Large language models (LLMs) have demonstrated impressive capabilities in natural language processing. However, their internal mechanisms are still unclear and this lack of transparency poses unwanted risks for downstream applications. Therefore, understanding and explaining these models is crucial for elucidating their behaviors, limitations, and social impacts. In this article, we introduce a taxonomy of explainability techniques and provide a structured overview of methods for explaining Transformer-based language models. We categorize techniques based on the training paradigms of LLMs: traditional fine-tuning-based paradigm and prompting-based paradigm. For each paradigm, we summarize the goals and dominant approaches for generating local explanations of individual predictions and global explanations of overall model knowledge. We also discuss metrics for evaluating generated explanations and discuss how explanations can be leveraged to debug models and improve performance. Lastly, we examine key challenges and emerging opportunities for explanation techniques in the era of LLMs in comparison to conventional deep learning models. [ABSTRACT FROM AUTHOR]

There has been a recent surge of interest in automating software engineering tasks using deep learning. This article addresses the problem of code generation, in which the goal is to generate target code given source code in a different language or a natural language description. Most state-of-the-art deep learning models for code generation use training strategies primarily designed for natural language. However, understanding and generating code requires a more rigorous comprehension of the code syntax and semantics. With this motivation, we develop an encoder-decoder Transformer model in which both the encoder and decoder are explicitly trained to recognize the syntax and dataflow in the source and target codes, respectively. We not only make the encoder structure aware by leveraging the source code's syntax tree and dataflow graph, but we also support the decoder in preserving the syntax and dataflow of the target code by introducing two novel auxiliary tasks: Abstract Syntax Tree (AST) path prediction and dataflow prediction. To the best of our knowledge, this is the first work to introduce a structure-aware Transformer decoder that models both syntax and dataflow to enhance the quality of generated code. The proposed StructCoder model achieves state-of-the-art performance on code translation and text-to-code generation tasks in the CodeXGLUE benchmark and improves over baselines of similar size on the APPS code generation benchmark. Our code is publicly available at https://github.com/reddy-lab-code-research/StructCoder/. [ABSTRACT FROM AUTHOR]

Recommending APIs is a practical and essential feature of IDEs. Improving the accuracy of API recommendations is an effective way to improve coding efficiency. With the success of deep learning in software engineering, the state-of-the-art (SOTA) performance of API recommendation is also achieved by deep-learning-based approaches. However, existing SOTAs either only consider the API sequences in the code snippets or rely on complex operations for extracting hand-crafted features, all of which have potential risks in under-encoding the input code snippets and further resulting in sub-optimal recommendation performance. To this end, this article proposes to utilize the code understanding ability of existing general code Pre-Training Models to fully encode the input code snippet to improve the accuracy of APIRecommendation, namely, PTM-APIRec. To ensure that the code semantics of the input are fully understood and the API recommended actually exists, we use separate vocabularies for the input code snippet and the APIs to be predicted. The experimental results on the JDK and Android datasets show that PTM-APIRec surpasses existing approaches. Besides, an effective way to improve the performance of PTM-APIRec is to enhance the pre-trained model with more pre-training data (which is easier to obtain than API recommendation datasets). [ABSTRACT FROM AUTHOR]

With the widespread success of deep learning technologies, many trained deep neural network (DNN) models are now publicly available. However, directly reusing the public DNN models for new tasks often fails due to mismatching functionality or performance. Inspired by the notion of modularization and composition in software reuse, we investigate the possibility of improving the reusability of DNN models in a more fine-grained manner. Specifically, we propose two modularization approaches named CNNSplitter and GradSplitter, which can decompose a trained convolutional neural network (CNN) model for N-class classification into N small reusable modules. Each module recognizes one of the N classes and contains a part of the convolution kernels of the trained CNN model. Then, the resulting modules can be reused to patch existing CNN models or build new CNN models through composition. The main difference between CNNSplitter and GradSplitter lies in their search methods: the former relies on a genetic algorithm to explore search space, while the latter utilizes a gradient-based search method. Our experiments with three representative CNNs on three widely used public datasets demonstrate the effectiveness of the proposed approaches. Compared with CNNSplitter, GradSplitter incurs less accuracy loss, produces much smaller modules (19.88% fewer kernels), and achieves better results on patching weak models. In particular, experiments on GradSplitter show that (1) by patching weak models, the average improvement in terms of precision, recall, and F1-score is 17.13%, 4.95%, and 11.47%, respectively, and (2) for a new task, compared with the models trained from scratch, reusing modules achieves similar accuracy (the average loss of accuracy is only 2.46%) without a costly training process. Our approaches provide a viable solution to the rapid development and improvement of CNN models. [ABSTRACT FROM AUTHOR]

Code summarization aims to generate short functional descriptions for source code to facilitate code comprehension. While Information Retrieval (IR) approaches that leverage similar code snippets and corresponding summaries have led the early research, Deep Learning (DL) approaches that use neural models to capture statistical properties between code and summaries are now mainstream. Although some preliminary studies suggest that IR approaches are more effective in some cases, it is currently unclear how effective the existing approaches can be in general, where and why IR/DL approaches perform better, and whether the integration of IR and DL can achieve better performance. Consequently, there is an urgent need for a comprehensive study of the IR and DL code summarization approaches to provide guidance for future development in this area. This article presents the first large-scale empirical study of 18 IR, DL, and hybrid code summarization approaches on five benchmark datasets. We extensively compare different types of approaches using automatic metrics, we conduct quantitative and qualitative analyses of where and why IR and DL approaches perform better, respectively, and we also study hybrid approaches for assessing the effectiveness of integrating IR and DL. The study shows that the performance of IR approaches should not be underestimated, that while DL models perform better in predicting tokens from method signatures and capturing structural similarities in code, simple IR approaches tend to perform better in the presence of code with high similarity or long reference summaries, and that existing hybrid approaches do not perform as well as individual approaches in their respective areas of strength. Based on our findings, we discuss future research directions for better code summarization. [ABSTRACT FROM AUTHOR]

Automated Program Repair (APR) is defined as the process of fixing a bug/defect in the source code, by an automated tool. APR tools have recently experienced promising results by leveraging state-of-the-art Neural Language Processing (NLP) techniques. APR tools such as TFix and CodeXGLUE that combine text-to-text transformers with software-specific techniques are outperforming alternatives, these days. However, in most APR studies, the train and test sets are chosen from the same set of projects (i.e., when APR fixes a bug in the test set from project A, the model has already seen example fixed bugs from project A in the training set). In the real world, however, APR models are meant to be generalizable to new and different projects. Therefore, there is a potential threat that reported APR models with high effectiveness perform poorly when the characteristics of the new project or its bugs are different than the training set's ("Domain Shift"). In this study, we first define the problem of domain shift in automated program repair. Next, we measure the potential damage of domain shift on two recent APR models (TFix and CodeXGLUE). Based on this observation, we then propose a domain adaptation framework that can adapt an APR model for a given target project. We conduct an empirical study with three domain adaptation methods FullFineTuning, TuningWithLightWeightAdapterLayers, and CurriculumLearning and two APR models on 2,672 bugs from 12 projects. The results show that our proposed framework on average can improve the effectiveness of TFix by 13.05% and CodeXGLUE by 48.78%, in terms of "Exact Match". Through experiments, we also show that the framework provides high efficiency and reliability (in terms of "Exposure Bias"). Using synthetic data to domain adapt TFix and CodeXGLUE on the projects with no data (Zero-shot learning), also results in an average improvement of 5.76% and 17.62% for TFix and CodeXGLUE, respectively. [ABSTRACT FROM AUTHOR]

In this article, we conduct a measurement study to comprehensively compare the accuracy impacts of multiple embedding options in cryptographic API completion tasks. Embedding is the process of automatically learning vector representations of program elements. Our measurement focuses on design choices of three important aspects, program analysis preprocessing, token-level embedding, and sequence-level embedding. Our findings show that program analysis is necessary even under advanced embedding. The results show 36.20% accuracy improvement, on average, when program analysis preprocessing is applied to transfer bytecode sequences into API dependence paths. With program analysis and the token-level embedding training, the embedding dep2vec improves the task accuracy from 55.80% to 92.04%. Moreover, only a slight accuracy advantage (0.55%, on average) is observed by training the expensive sequence-level embedding compared with the token-level embedding. Our experiments also suggest the differences made by the data. In the cross-app learning setup and a data scarcity scenario, sequence-level embedding is more necessary and results in a more obvious accuracy improvement (5.10%). [ABSTRACT FROM AUTHOR]

Automated program repair (APR) aims to fix software bugs automatically and plays a crucial role in software development and maintenance. With the recent advances in deep learning (DL), an increasing number of APR techniques have been proposed to leverage neural networks to learn bug-fixing patterns from massive open-source code repositories. Such learning-based techniques usually treat APR as a neural machine translation (NMT) task, where buggy code snippets (i.e., source language) are translated into fixed code snippets (i.e., target language) automatically. Benefiting from the powerful capability of DL to learn hidden relationships from previous bug-fixing datasets, learning-based APR techniques have achieved remarkable performance. In this article, we provide a systematic survey to summarize the current state-of-the-art research in the learning-based APR community. We illustrate the general workflow of learning-based APR techniques and detail the crucial components, including fault localization, patch generation, patch ranking, patch validation, and patch correctness phases. We then discuss the widely adopted datasets and evaluation metrics and outline existing empirical studies. We discuss several critical aspects of learning-based APR techniques, such as repair domains, industrial deployment, and the open science issue. We highlight several practical guidelines on applying DL techniques for future APR studies, such as exploring explainable patch generation and utilizing code features. Overall, our article can help researchers gain a comprehensive understanding about the achievements of the existing learning-based APR techniques and promote the practical application of these techniques. Our artifacts are publicly available at the repository: https://github.com/iSEngLab/AwesomeLearningAPR. [ABSTRACT FROM AUTHOR]

Machine learning (ML) has been increasingly used in a variety of domains, while solving ML programming tasks poses unique challenges due to the fundamental difference in the nature and the construct of general programming tasks, especially for developers who do not have ML backgrounds. Automatic code generation that produces a code snippet from a natural language description can be a promising technique to accelerate ML programming tasks. In recent years, although many deep learning-based neural code generation models have been proposed with high accuracy, the fact that most of them are mainly evaluated on general programming tasks calls into question their effectiveness and usefulness in ML programming tasks. In this article, we set out to investigate the effectiveness of existing neural code generation models on ML programming tasks. For our analysis, we select six state-of-the-art neural code generation models and evaluate their performance on four widely used ML libraries, with newly created 83K pairs of natural-language described ML programming tasks. Our empirical study reveals some good, bad, and missing aspects of neural code generation models on ML tasks, with a few major ones listed below. (Good) Neural code generation models perform significantly better on ML tasks than on non-ML tasks with an average difference of 10.6 points in BLEU-4 scores. (Bad) More than 80% of the generated code is semantically incorrect. (Bad) Code generation models do not have significance in improving developers' completion time. (Good) The generated code can help developers write correct code by providing developers with clues for using correct APIs. (Missing) The observation from our user study reveals the missing aspects of code generation for ML tasks, e.g., decomposing code generation for divide-and-conquer into API sequence identification and API usage generation. [ABSTRACT FROM AUTHOR]

Code writing is repetitive and predictable, inspiring us to develop various code intelligence techniques. This survey focuses on code search, that is, to retrieve code that matches a given natural language query by effectively capturing the semantic similarity between the query and code. Deep learning, being able to extract complex semantics information, has achieved great success in this field. Recently, various deep learning methods, such as graph neural networks and pretraining models, have been applied to code search with significant progress. Deep learning is now the leading paradigm for code search. In this survey, we provide a comprehensive overview of deep learning-based code search. We review the existing deep learning-based code search framework that maps query/code to vectors and measures their similarity. Furthermore, we propose a new taxonomy to illustrate the state-of-the-art deep learning-based code search in a three-step process: query semantics modeling, code semantics modeling, and matching modeling, which involves the deep learning model training. Finally, we suggest potential avenues for future research in this promising field. [ABSTRACT FROM AUTHOR]

Cross-domain recommender systems could potentially improve the recommendation performance by means of transferring abundant knowledge from the auxiliary domain to the target domain. They could help address some key challenges in recommender systems, such as data sparsity and cold start. However, most existing cross-domain recommendation approaches represent the user preferences based on a single kind of user's feature or behavior and fail to explore the hidden interaction effects of different kinds of features or behaviors. In this article, we propose the Sequential and Graphical Cross-Domain Recommendations with a Multi-View Hierarchical Transfer Gate (SGCross) to transfer user representations from multiple perspectives. The SGCross model constructs a user profile by learning the personal preference from a personal view, the dynamic preference from a temporal view, as well as the collaborative preference from a collaborative view. Specifically, a Multi-view Hierarchical Gate (MHG) is designed to transfer the informative representations of user knowledge on different views from the auxiliary domain separately, aiming to enhance the user representations. Furthermore, a two-stage attentive fusion module is designed to integrate transferred information at two levels: the domain level and the view level. Extensive experiments on the Amazon dataset and the Douban dataset have demonstrated that SGCross effectively improves the accuracy of cross-domain recommendations and outperforms the state-of-the-art baseline models. [ABSTRACT FROM AUTHOR]

This survey documents representation approaches for classification across different modalities, from purely content-based methods to techniques utilizing external sources of structured knowledge. We present studies related to three paradigms used for representation, namely (a) low-level template-matching methods, (b) aggregation-based approaches, and (c) deep representation learning systems. We then describe existing resources of structure knowledge and elaborate on the need for enriching representations with such information. Approaches that utilize knowledge resources are presented next, organized with respect to how external information is exploited, i.e., (a) input enrichment and modification, (b) knowledge-based refinement and (c) end-to-end knowledge-aware systems. We subsequently provide a high-level discussion to summarize and compare strengths/weaknesses of the representation/enrichment paradigms proposed, and conclude the survey with an overview of relevant research findings and possible directions for future work. [ABSTRACT FROM AUTHOR]

Progress in natural language processing has been dictated by the rule of more: more data, more computing power, more complexity, best exemplified by deep learning Transformers. However, training (or fine-tuning) large dense models for specific applications usually requires significant amounts of computing resources. One way to ameliorate this problem is through data engineering instead of the algorithmic or hardware perspectives. Our focus here is an under-investigated data engineering technique, with enormous potential in the current scenario - Instance Selection (IS) (a.k.a. Selective Sampling, Prototype Selection). The IS goal is to reduce the training set size by removing noisy or redundant instances while maintaining or improving the effectiveness (accuracy) of the trained models and reducing the training process cost. We survey classical and recent state-of-the-art IS techniques and provide a scientifically sound comparison of IS methods applied to an essential natural language processing task--Automatic Text Classification (ATC). IS methods have been normally applied to small tabular datasets and have not been systematically compared in ATC. We consider several neural and non-neural state-of-the-art ATC solutions and many datasets. We answer several research questions based on tradeoffs induced by a tripod (training set reduction, effectiveness, and efficiency). Our answers reveal an enormous unfulfilled potential for IS solutions. Specially, we show that in 12 out of 19 datasets, specific IS methods--namely, Condensed Nearest Neighbor, Local Set-based Smoother, and Local Set Border Selector--can reduce the size of the training set without effectiveness losses. Furthermore, in the case of fine-tuning the Transformer methods, the IS methods reduce the amount of data needed, without losing effectiveness and with considerable training-time gains. [ABSTRACT FROM AUTHOR]

The remarkable success of deep learning in various domains relies on the availability of large-scale annotated datasets. However, obtaining annotations is expensive and requires great effort, which is especially challenging for videos. Moreover, the use of human-generated annotations leads to models with biased learning and poor domain generalization and robustness. As an alternative, self-supervised learning provides a way for representation learning that does not require annotations and has shown promise in both image and video domains. In contrast to the image domain, learning video representations are more challenging due to the temporal dimension, bringing in motion and other environmental dynamics. This also provides opportunities for video-exclusive ideas that advance self-supervised learning in the video and multimodal domains. In this survey, we provide a review of existing approaches on self-supervised learning focusing on the video domain. We summarize these methods into four different categories based on their learning objectives: (1) pretext tasks, (2) generative learning, (3) contrastive learning, and (4) cross-modal agreement. We further introduce the commonly used datasets, downstream evaluation tasks, insights into the limitations of existing works, and the potential future directions in this area. [ABSTRACT FROM AUTHOR]

Neuromorphic Computing, a concept pioneered in the late 1980s, is receiving a lot of attention lately due to its promise of reducing the computational energy, latency, as well as learning complexity in artificial neural networks. Taking inspiration from neuroscience, this interdisciplinary field performs a multi-stack optimization across devices, circuits, and algorithms by providing an end-to-end approach to achieving brain-like efficiency in machine intelligence. On one side, neuromorphic computing introduces a new algorithmic paradigm, known as Spiking Neural Networks (SNNs), which is a significant shift from standard deep learning and transmits information as spikes (“1” or “0”) rather than analog values. This has opened up novel algorithmic research directions to formulate methods to represent data in spike-trains, develop neuron models that can process information over time, design learning algorithms for event-driven dynamical systems, and engineer network architectures amenable to sparse, asynchronous, event-driven computing to achieve lower power consumption. On the other side, a parallel research thrust focuses on development of efficient computing platforms for new algorithms. Standard accelerators that are amenable to deep learning workloads are not particularly suitable to handle processing across multiple timesteps efficiently. To that effect, researchers have designed neuromorphic hardware that rely on event-driven sparse computations as well as efficient matrix operations. While most large-scale neuromorphic systems have been explored based on CMOS technology, recently, Non-Volatile Memory (NVM) technologies show promise toward implementing bio-mimetic functionalities on single devices. In this article, we outline several strides that neuromorphic computing based on spiking neural networks (SNNs) has taken over the recent past, and we present our outlook on the challenges that this field needs to overcome to make the bio-plausibility route a successful one. [ABSTRACT FROM AUTHOR]