Heterogeneous defect prediction (HDP) refers to predicting defect-proneness of software modules in a target project using heterogeneous metric data from other projects. Existing HDP methods mainly focus on instances with single source. In practice, there exist plenty external Multiple sources can generally provide more information than project. Therefore, it is meaningful investigate whether the performance be improved by employing multiple sources. However, precondition conducting that are...

We present a novel Automatic Speech Recognition (ASR) dataset for the Oromo language, widely spoken language in Ethiopia and neighboring regions. The was collected through crowd-sourcing initiative, encompassing diverse range of speakers phonetic variations. It consists 100 hours real-world audio recordings paired with transcriptions, covering read speech both clean noisy environments. This addresses critical need ASR resources which is underrepresented. To show its applicability task, we...

Environmental perception is a critical component of automated driving systems. Advancing environmental algorithms toward applications in open-world road scenarios current research trend. However, traditional single-modality detectors exhibit weak generalization capabilities, while modern vision-language models require manual input text prompts to function properly. Since conventional methods fail meet the requirements perception, this study proposes prompt-free model address problem. The...

Software defect prediction is an important software quality assurance technique. Nevertheless, the performance of constructed model easily susceptible to irrelevant or redundant features in projects and not predominant enough. To address these two issues, a novel called SSEPG based on Stacked Sparse Denoising AutoEncoders (SSDAE) Extreme Learning Maching (ELM) optimised by Particle Swarm Optimisation (PSO) another complementary Gravitational Search Algorithm (GSA) are proposed this paper,...

A Lift Dispatching System (LDS) is a typical real-time system that highly complicated in design and implementation. This article presents the formal design, specification, modeling of LDS using denotational mathematics known as Real-Time Process Algebra (RTPA). The conceptual model introduced initial requirements for system. architectural created RTPA methodologies refined by set Unified Data Models (UDMs). static behaviors are specified (UPMs) lift dispatching serving processes. dynamic...

The study on the speech emotion recognition has very important realistic values in such aspects as enhancing intelligence and humanity of computer, developing new human-machine environment improving results. first goal is to search most useful features with analyzing related emotions. second gold find a model make use these features. basic course introduced, which includes signal preprocess feature extraction recognition. After choosing Mel-Frequency Cepstral Coefficients (MFCC) its...

Software defect prediction plays an important role in software quality assurance. However, the performance of model is susceptible to irrelevant and redundant features. In addition, previous studies mostly regard as a single objective optimization problem, multi-objective has not been thoroughly investigated. For above two reasons, we propose following solutions this paper: (1) leverage advanced deep neural network—Stacked Contractive AutoEncoder (SCAE) extract robust semantic features from...

Reusing software architecture and its constituents, which are a kind of coarse-grained resources, confronted with great difficulty. Reflection mechanism has been successfully applied in the reuse code component, but scarcely constituents. Architectural reflection is computation performed by system about own architecture, currently confined realm runtime management dynamic evolution systems. This paper proposes supporting architectural level design, offers information support for reusing...

Software defect prediction is a research hotspot in the field of software engineering. However, due to limitations current machine learning algorithms, we can’t achieve good effect for by only using algorithms. In previous studies, some researchers used extreme (ELM) conduct prediction. initial weights and biases ELM are determined randomly, which reduces performance ELM. Motivated idea search based engineering, propose novel model named KAEA on kernel principal component analysis (KPCA),...

Various informative factors mixed in speech signals, leading to great difficulty when decoding any of the factors. An intuitive idea is factorize each frame into individual factors, though it turns out be highly difficult. Recently, we found that speaker traits, which were assumed long-term distributional properties, are actually short-time patterns, and can learned by a carefully designed deep neural network (DNN). This discovery motivated cascade factorization (CDF) framework will...

The traditional Architectural Description Languages ADLs lack the ability to describe crosscutting concerns and interactions in software architecture, which result design of system difficulty comprehension, evolution reuse. This paper proposes a new Aspect-Oriented ADL-AC2-ADL provide formal basis for representation tangling scattering establish architecture with higher dependability.

Software architecture design is an important stage in the whole software development lifecycle. Yet non-functional requirements (NFRs) have been afflicted architects for a long time and how to efficiently address them also becomes big headache problem. This paper focuses on NFRs offers preliminary insight separating multiple dimensional way. Consequently, it proposes model named "1+X" handling architecture, including classifying those into 12 dimensions, then specifying mapping aspects. By...

Crosscutting behaviors and features of architectural units have always been a tricky issue in software architecture design. If not well treated, they may cause some unnecessary coupling among hamper maintenance, evolution reusability products. Appropriate modeling approaches expressions contribute to the solution these problems. This paper proposes an Aspect-Oriented Software Architecture Design approach AOSAD design systems. employs special kind component called aspectual encapsulate...

Constraint solution reuse is an effective approach to save the time of constraint solving in symbolic execution. Most existing approaches are based on syntactic or semantic equivalence constraints; e.g. Green framework able constraints which have different representations but semantically equivalent, through canonizing into syntactically equivalent normal forms. However, syntactic/semantic not a necessary condition for reuse--some their solutions still potential reuse. Existing unable...