The proceedings contain 44 papers. The topics discussed include: text analysis method based on multi-channel parallel classifier;a rudimentary proof on Goldbach conjectures;a combined algorithm for imbalanced classifi...
ISBN:
(纸本)9781665482202
The proceedings contain 44 papers. The topics discussed include: text analysis method based on multi-channel parallel classifier;a rudimentary proof on Goldbach conjectures;a combined algorithm for imbalanced classification based on dual distribution representation learning and classifier decoupling learning;transformer-based deep learning method for the prediction of ventilator pressure;multiple input single target streamflow forecast by neurowavelet networks;STDE: a single-senior-teacher knowledge distillation model for high-dimensional knowledge graph embeddings;a survey: complex knowledge base question answering;neural data-to-text generation guided by predicted plan;design and implementation of a perioperative medical data quality management platform;multi-view user preference learning with knowledge graph for recommendation;and representation learning of knowledge graph integrating entity description and language morphological structure information.
Closing the loop of the traditional linear economic model of take-make-dispose is primarily facilitated by integrating reverse logistics supply chains (RLSC) into the forward supply chain in the construction industry....
详细信息
The proceedings contain 67 papers. The topics discussed include: restricted metamodel-based similarity propagation: a comparative study;a collaborative approach to capture the domain language;variable-based analysis f...
ISBN:
(纸本)9789972825804
The proceedings contain 67 papers. The topics discussed include: restricted metamodel-based similarity propagation: a comparative study;a collaborative approach to capture the domain language;variable-based analysis for traceability in QVT-R model transformations;improving quality models construction through knowledge reuse;verification of software process line models: a checklist-based inspection approach;providing software maintenance and evolution as a service in a small organization: an approach based on CMMI-DEV and CMMI-SVC;a practical experience of a software process line creation;methodologies for evaluation and improvement of software processes in the context of quality and maturity models: a systematic mapping;an evaluation of functional size measurement methods;mechanisms to characterize context of empirical studies in softwareengineering;lightweight software verification with pluggable type-checking;and a computational infrastructure for research synthesis in softwareengineering.
Modern Code Review (MCR) is an essential practice in softwareengineering, supporting early defect detection, enhancing code quality, and fostering knowledge. To manage code review tasks effectively, developers need t...
详细信息
ISBN:
(数字)9798331535100
ISBN:
(纸本)9798331535117
Modern Code Review (MCR) is an essential practice in softwareengineering, supporting early defect detection, enhancing code quality, and fostering knowledge. To manage code review tasks effectively, developers need to understand the intent behind code changes, such as a bug fix, test, refactoring, or new feature. Traditional methods for categorizing code changes in MCR rely on rule-based heuristics with predefined keywords. However, these methods lack context regarding the code changes, leading to limited generalizability, particularly when dealing with sparsely documented changes. This paper addresses these limitations by investigating the potential of Large Language Models (LLMs) for changes' intent classification. We introduce LLM Change Classifier (LLMCC), an LLM-based approach that classifies code changes based on their underlying intent. We evaluate the effectiveness of LLMCC by conducting an empirical study on three open-source projects: Android, OpenS tack, and Qt. The performance of LLMCC was benchmarked against traditional heuristic methods, conventional machine learning algorithms (including Decision Trees and Random Forests), and state-of-the-art transformer models (including BERT and RoBERTa). Results show that LLMCC significantly enhances code change intent classification accuracy, achieving up to a 33 % improvement in F1 score over heuristic-based methods. Additionally, LLMCC outperformed both traditional machine learning and transformer models, achieving an average 77% improvement in terms of Matthew Correlation Coefficient (MCC). These findings underscore the potential of LLMCC to streamline code change intent classification.
Logs record essential information about system operations and serve as a critical source for anomaly detection, which has generated growing research interest. Utilizing large language models (LLMs) within a retrieval-...
详细信息
ISBN:
(数字)9798350368741
ISBN:
(纸本)9798350368758
Logs record essential information about system operations and serve as a critical source for anomaly detection, which has generated growing research interest. Utilizing large language models (LLMs) within a retrieval-augmented generation (RAG) framework for log-based anomaly detection is an effective approach due to its strong generalization capabilities and efficient few-shot performance. However, the effectiveness of this method hinges on the quality of the knowledge source, which can be impacted by noise and changes within the software systems. Facing these problems, in this paper, we propose a novel log-based anomaly detection method named EagerLog, employing active learning to choose the logs for humans to label, thereby adding them to the knowledge source, thus enhancing the knowledge source and maintaining its quality. Our experiments on three open datasets (BGL, Thunderbird, Zookeeper) and one industrial dataset demonstrate that EagerLog can achieve 93.65% F1 score with approximately 10 labeled log sequences, surpassing existing methods by 15.32%.
Over recent years there has been an increase in the use of generic Computational Fluid Dynamics (CFD) software packages spread across various application fields. This has created the need for the integration of expert...
详细信息
Over recent years there has been an increase in the use of generic Computational Fluid Dynamics (CFD) software packages spread across various application fields. This has created the need for the integration of expertise into CFD software. Expertise can be integrated into CFD software in the form of an Intelligent knowledge-based System (IKBS). The advantages of integrating intelligence into generic engineeringsoftware are discussed with a special view to softwareengineering considerations. The software modelling cycle of a typical engineering problem is identified and the respective expertise and user control needed for each modelling phase is shown. The requirements of an IKBS for CFD software are discussed and compared to current practice. The blackboard software architecture is presented. This is shown to be appropriate for the integration of an IKBS into an engineeringsoftware package. This is demonstrated through the presentation of the prototype CFD software package FLOWES. Copyright (C) 1996 Civil-Comp Limited and Elsevier Science Limited.
The proceedings contains 144 papers from the Seventh IASTED International conference on softwareengineering and Applications. The topics discussed include: the component model for web services;mobile computer systems...
详细信息
ISBN:
(纸本)0889863946
The proceedings contains 144 papers from the Seventh IASTED International conference on softwareengineering and Applications. The topics discussed include: the component model for web services;mobile computer systems for automated analysis of ABR recordings;enhancing the power of OLAP with knowledge discovery;quantitative metrics for risk assessment in software projects;a dialog flow notation for web-based applications;and issues in analyzing dynamic system evolution.
The proceedings contain 11 papers. The topics discussed include: assessment of a framework for comparing software architecture analysis methods;systematic review of statistical process control: an experience report;pr...
The proceedings contain 11 papers. The topics discussed include: assessment of a framework for comparing software architecture analysis methods;systematic review of statistical process control: an experience report;preliminary results of a study of the completeness and clarity of structured abstracts;an experiment measuring the effects of maintenance tasks on program knowledge;outsourcing and knowledge management in software testing;predicting web development effort using a Bayesian network;motivators of software process improvement: an analysis of Vietnamese practitioners’ views;and experimental comparison of the comprehensibility of a UML-based formal specification versus a textual one.
The proceedings contain 72 papers. The topics discussed include: construction and application of knowledge graph for food therapy;determining the most significant metadata features to indicate defective software commi...
ISBN:
(纸本)9798350345889
The proceedings contain 72 papers. The topics discussed include: construction and application of knowledge graph for food therapy;determining the most significant metadata features to indicate defective software commits;documentation practices in agile software development: a systematic literature review;rule-based representation learning for traditional Chinese medicine knowledge graph;experimental evaluation of adversarial attacks against natural language machine learning models;mobile user analysis considering collaboration with financial services;commit message can help: security patch detection in open source software via transformer;texture and orientation-based feature extraction for robust facial expression recognition;scientific organization of blood donation camp through lexicographic optimization and taxicab path computation;research and application of content-based image hash retrieval algorithm;and cloud-based digital twins storage in emergency healthcare.
knowledge distillation has become a crucial technique for transferring intricate knowledge from a teacher model to a smaller student model. While logit-basedknowledge distillation has shown promise, existing methods ...
详细信息
ISBN:
(数字)9798350368741
ISBN:
(纸本)9798350368758
knowledge distillation has become a crucial technique for transferring intricate knowledge from a teacher model to a smaller student model. While logit-basedknowledge distillation has shown promise, existing methods often overlook the efficient distillation of logits. In this paper, we introduce a novel approach called Class-wise Adaptive Logits Distillation (CALD) based on meta-learning. Our method leverages a meta-network to generate class-adaptive weights, delivering both explicit and implicit knowledge adaptively. By training the meta-network to assign higher weights to specific classes crucial for the student model’s learning from the teacher model, our approach enhances the knowledge transfer process. Experimental results on CIFAR-100 and ImageNet datasets demonstrate that CALD surpasses state-of-the-art knowledge distillation methods, achieving enhanced accuracy and efficiency in transferring knowledge from teacher to student models.
暂无评论