Machine learning (ML) models in production fail when their broader systems – from data pipelines to deployment environments – deviate from training assumptions, not merely due to statistical anomalies in input data. Despite extensive work on data drift, data validation, and out-of-distribution detection, ML monitoring research remains largely model-centric while neglectingcontextual information: auxiliary signals about the system around the model (external factors, data pipelines, downstream applications). Incorporating this context turns statistical anomalies into actionable alerts and structured root-cause analysis. Drawing on a systematic review of 94 primary studies, we identify three dimensions of contextual information for ML monitoring: thesystemelement concerned (natural environment or technical infrastructure); theaspectof that element (runtime states, structural relationships, prescriptive properties); and therepresentationused (formal constructs or informal formats). This forms theContextual System-Aspect-Representation(C-SAR) framework, a descriptive model synthesizing our findings. We identify 20 recurring triplets across these dimensions and map them to the monitoring activities they support. This study provides a holistic perspective on ML monitoring: from interpreting “tea leaves” (i.e., isolated data and performance statistics) to constructing and managing “system maps” (i.e., end-to-end views that connect data, models, and operating context).

In the last two decades, research in Self-Adaptive Systems (SAS) has proposed various approaches for inducing a software system with the ability to change itself at runtime in terms of self-adaptation strategies. For the wider adoption of these strategies, there is a need for a framework and tool support to enable their analysis, evaluation, comparison, and eventually their selection in overlapping cases. In this paper, we take a step in this direction by proposing a comprehensive metric suite, i.e., the Adaptive Strategies Metric Suite (ASMS), to measure the design and runtime properties of the adaptive strategies for SAS. ASMS consists of metrics that can be applied through both static and dynamic code analysis. The metrics pertaining to static code analysis have been implemented as a plugin for Understand tool.

Self-Adaptive Systems are usually built of a managed part, implementing their functionality, and a managing part, implementing their self-adaptation. The complexity of self-adaptive systems results also from the existence of the managing part and the interaction between the managed and the managing parts. The nonself- adaptive systems may be seen as the managed part of self-adaptive systems. The self-adaptive systems are evaluated based on their performances resulted from the self-adaptation. However, self-adaptive systems are software systems, hence, also their software quality is equally important. Our analysis compares the internal quality of self-adaptive and non-self-adaptive systems by considering code smells, architectural smells, and GoF's design patterns. This comparison provides an insight to the health of the self-adaptive systems with respect to the non-self-adaptive systems (the last being considered as a quality reference).

Software development is more and more influenced by the usage of FLOSS (Free, Libre and Open Source Software) projects. These software projects are developed in web collaborative environments hosted on web platforms, called code forges. Many code forges exist, with different capabilities. GitHub is perhaps the largest code forge available, and many projects have been migrated from different code forges to GitHub. Given its success, we want to understand if its adoption has effect on the projects' internal quality. To consider objective measures of internal quality, we apply four known tools performing static analysis to extract metrics and code anomalies. These data is extracted on six versions of six FLOSS projects, and compared to understand if the migration to GitHub had any consistent effect over any of the considered measures.

GitHub is a platform used for the development of software projects. It provides a traceable project repository and a social meeting place for communities of practices. This poster presents the students' feedback on using GitHub as a development platform for software projects counting as an exam for a 3rd-year undergraduate software engineering course on software design. Students worked in teams and their feedback is positive overall.

The safety of the persons in work environments is an open issue. The diffusion of smart solutions including Smart Work Environments has lead to new and enhanced approaches which consider the dynamic nature of such environments. Smart solutions use context information and adapt themselves based on the current context and needs at run-time. However, as traditional approaches have been proven to be effective, they should be extended and adapted to run-time safety management by focusing on the dynamic nature of smart environments. In this work, we propose transforming traditional safety management approaches into models that can be updated and used at run-time.

To better understand and exploit the knowledge necessary to comprehend and evolve an existing system, different models can be extracted from it. Models represent the extracted information at various abstraction levels, and are useful to document, maintain, and reengineer the system. The Knowledge Discovery Metamodel (KDM) has been defined by the object management group as a meta-model supporting a large share of reverse engineering activities. Its specification has also been adopted by the ISO in 2012. This paper explores and describes alternative meta-models proposed in the literature to support reverse engineering, program comprehension, and software evolution activities. We focus on the similarity and differences of the alternative meta-models with KDM, trying to understand the potentials of reciprocal information interchange. We describe KDM and other five meta-models, plus their extensions available in the literature and their diffusion in the reverse engineering community. We also investigate the approaches using KDM and the five meta-models. In the paper, we underline the limited reuse of models for reverse engineering, and identify potential directions for future related research, to enhance the existing models and ease the exchange of information among them.

Self-protection mechanisms aim to improve security of software systems at runtime. They are able to automatically prevent and/or react to security threats by observing the state of a system and its execution environment, by reasoning on the observed state, and by applying enhanced security strategies appropriate for the current threat. Self-protection mechanisms complement traditional security solutions which are mostly static and focus on the boundaries of a system, missing in this way the overall picture of a system's security. This paper presents several self-protection mechanisms which have been developed in the context of a case study concerning a home banking system. Essentially, the mechanisms described in this paper aim to improve the security of the system in the following two scenarios: users' login and bank operations. Furthermore, the proposed self-protection mechanisms are presented through the taxonomy proposed in (Yuan, 2014).

Architecture erosion constitutes the most visible effect of the degradation of design. It is a major reason to address the design debt, often caused by architectural mismatch problems. Today, the identification of design erosion is a major concern for designers and software maintainers. Adequate tools are necessary to identify and repair the debt. This paper describes our experience on identifying architectural erosion problems on existing open source software applications through the support of two well known tools. Moreover, we outline if the tools provide useful hints in repairing some of the detected problems.

This paper describes our joint experience in organizing a Doctoral Symposium co-located with one of the main software engineering conferences. It presents the issues we addressed during the organization of the symposium, as well as during the symposium. This paper is addressed (1) to the organizers of Doctoral Symposiums aiming to provide a recipe with the main ingredients and preparation steps and their related significance in the entire organization process, as well as (2) to the PhD students providing our feedback on the expectations, evaluation, and presentation of their contributions.

SUMMARY Design patterns play a significant role in reverse engineering by providing information not only on how but also on why a solution has been implemented in a specific way because of their semantics. The application of design patterns leads to their personalization to a specific context, hence to the generation of variants. This makes their recognition a challenging task, which may be addressed through the understanding and detection of the micro‐structures design patterns are made of. This is very useful for the detection as well as for the application of design patterns. The principal aim of this paper is to present a survey on these micro‐structures and a comparison among them in the perspective of reverse engineering. Because of their less complex structure and behavior, as well as closer link to the source code, the recognition of these micro‐structures may be automated, which can be considered a step towards the automatic recognition of the more complex design patterns. In this paper, we consider four of the most significant types of micro‐structures: elemental design patterns, clues, sub‐patterns, and micro patterns. To analyze the role of the micro‐structures in the design pattern detection process, we make a comparison among these four types of micro‐structures and among the micro‐structures of various types in order to identify the relations among them. Copyright © 2011 John Wiley & Sons, Ltd.

SUMMARY The paper proposes a characterization of risks and a service‐oriented prototype to face risky situations in work environments, such as in industrial plants or building construction areas. A risk is the overture of emergencies that produce human and/or material damages. Therefore, it is particularly critical to identify and manage risks to avoid their evolution into emergencies. In this paper, we outline the technological features of a risk environment and propose a risk model and a service‐based simulation prototype aimed to improve safety in work environments. We discuss engineering issues concerning risk modeling and management. Furthermore, we propose a risk management system solution and its related implemented prototype composed of services able to detect and also to prevent the occurrence of risk conditions. Copyright © 2012 John Wiley & Sons, Ltd.

Nowadays, mobile devices have an increasing role in the generation and visualization of multimedia content. The issues raised by the management of multimedia content in mobile environments are mostly due to the variety of available devices, their limitations in front of the traditional ones, the network characteristics, the variety of the multimedia content types and formats, as well as to the lack of common standards related to the provisioning of services. In this context, Alembik aims to address these limitations and to provide a framework to transcode and adapt the multimedia content at runtime to the properties of the requiring mobile device. The main keywords which characterize Alembik are: dynamicity, flexibility, extensibility, standardization, user transparency, platform independence, high performance, and open source.

Currently, research on reverse engineering and automated design pattern detection focuses mostly on some of the programming languages such as Java and C/C++, with marginal interest in the .NET area. In this paper, we present a tool for analyzing .NET executables for architecture reconstruction in general and design pattern detection in particular. The tool extracts the meaningful information from .NET systems and produces an output compatible with the design pattern detection tool we are developing for Java software, tool called MARPLE (Metrics and Architecture Reconstruction PLug-in for Eclipse).

Mobile widgets represent applications exploiting web technologies and providing specific functionalities in an efficient and user-friendly way. This paper presents our approach to the automatic generation of widgets, which is based on the separation of concerns between the specification of their structural and functional characteristics, and their appearance. The validation of our solution has been done through various case studies, among which we mention the DISCo widget, a mobile widget which provides academic information for the students of the Computer Science Department at the University of Milano-Bicocca in Italy.

Mobile devices, such as smart phones, PDAs, or laptops are part of our everyday life. They became ubiquitous and pervasive. Their benefits are exploited in more and more scenarios and application domains. This paper presents an example of how mobile devices and their related applications may be successfully used by the bike fans. The issue is to provide an application for all types of bike users (e.g., amateurs and professionals), application which is able to adapt dynamically to the exigencies of the users. Part of the functionalities is already provided by different applications on various devices (e.g., computation of a route plan, visualization of the current status of the users' health, visualization of forecast or tourist information) and we aim to integrate them into our solution, while part consists in enhanced and novel functionalities specific to the cyclists (e.g., computation and visualization of the altimetry of a route plan, visualization of information specific to a bike competition regarding the current user).

In this paper, we propose an approach to face emergencies and risky situations through real-time wearable services in industrial environments. We outline the social and technological features of a wearable-based service system, which aims to improve safety in workplaces. To achieve this goal, we propose a system composed of real-time wearable services constituting a platform able to prevent the occurrence of emergency and/or unsafe conditions in work contexts. After the presentation of the related work, the paper introduces the most significant social and technological aspects of the risk management. Furthermore, attention is focused on the modeling guidelines we have identified for the architectural approach which exploits the core concepts of the service-oriented paradigm.

The available frameworks for adaptive systems propose various solutions for addressing and implementing software which is adaptable at runtime. These solutions have different perspective on adaptivity: architectural, structural, or behavioral. Hence, based on which criteria should be decided which framework to adopt? In this paper, we present a possible approach to evaluate the available frameworks through a set of metrics which aim to capture the main issues of adaptive systems from the development perspective.

Java PathFinder (JPF) is an explicit state model checker developed by the Automated Software Engineering Group of NASA of the AMES Research Center (California). Eclipse is probably the most important and used Java integrated developing environment (IDE) and not only; it is a framework/environment that can be easily extended with new functionalities by exploiting its plug-in mechanism. Through a JPF plug-in for Eclipse it is possible to integrate powerful model checking and testing capabilities into the development environment. This paper presents the re-design of the standalone version of JPF towards an Eclipse plug-in exploiting and outlining in this way the advantages of an open source development