SE /

Seminar

SE Seminar Schedule

This is the schedule for the SE division seminar in the academic year 2020 / 2021. Common types of seminars are trial talks (for upcoming conferences, defenses, etc.), presentations of ongoing research, lectures by academic guests, and other research talks.

Date & TimePresenterTitleTalk Type Talk AbstractLocation
05.11.2021 (10:00)Khan Mohammad HabibullahNon-functional Requirements for Machine Learning: Understanding Current Use and Challenges in IndustryPost conference, and current research on NFRs for MLMachine Learning (ML) is an application of Artificial Intelligence (AI) that uses big data to produce complex predictions and decision-making systems, which would be challenging to obtain otherwise. To ensure the success of ML-enabled systems, it is essential to be aware of certain qualities of ML solutions (performance, transparency, fairness), known from a Requirement Engineering (RE) perspective as non-functional requirements (NFRs). However, when systems involve ML, NFRs for traditional software may not apply in the same ways; some NFRs may become more prominent or less important; NFRs may be defined over the ML model, data, or the entire system; and NFRs for ML may be measured differently. In this work, we aim to understand the state-of-the-art and challenges of dealing with NFRs for ML in industry. We interviewed ten engineering practitioners working with NFRs and ML. We find examples of (1) the identification and measurement of NFRs for ML, (2) identification of more and less important NFRs for ML, and (3) the challenges associated with NFRs and ML in the industry. This knowledge paints a picture of how ML-related NFRs are treated in practice and helps to guide future RE for ML efforts.Hybrid: Jupiter 473 and https://chalmers.zoom.us/j/68106313512 Password: 067177
05.11.2021 (10:45)Christoph Laaber (Simula, Norway)Variability of Microbenchmark Results and How to Deal with ItExternal VisitorPerformance variability is a well-known challenge in software systems research and in performance engineering practice. Microbenchmarks, a form of performance testing technique, is not immune to performance variability, which can lead to unreliable results from which one can not draw reliable conclusions. The reasons for performance variability are many-fold, e.g., the environment the benchmark is executed in, the way the benchmark is written, or the measurement methodology that is applied. In this talk, I will show empirical evidence on the extent of benchmark result variability and elaborate on two techniques that can help dealing with it.Hybrid: Jupiter 473 and https://chalmers.zoom.us/j/68106313512 Password: 067177
19.11.2021 (10:00)Hamdy Michael AyasThe journey of migrating towards microservicesConference presentation practiceWe will showcase the evolutionary and iterative nature of the migration journey towards microservices at an architectural-level and system-implementation level. Also, we identify 18 detailed activities that take place in these levels, categorized in the four phases of 1) designing the architecture, 2) altering the system, 3) setting up supporting artifacts, and 4) implementing additional technical artifacts.Hybrid: Jupiter 473 and https://chalmers.zoom.us/j/65639661065 Password: 701684
10.12.2022 (13:00)Ricardo Diniz CaldasEngineering Software for Resilient Cyber-Physical SystemsLicentiate Dry-RunResilient cyber-physical systems (CPS) should avoid, withstand, recover from, and evolve and adapt to cope with adversity stemming from computation, networking, or physical environment. From the engineering point of view, the usefulness of such systems is hindered by their lack of ability to adapt and overcome unknown stimuli, ever-changing and conflicting objectives, and deprecated internal components. Software as a tool for self-management is a key instrument for dealing with uncertainty. In this presentation, I discuss design, verification and validation of resilient CPS from software viewpoint and its implications. For more information, click here.Jupiter 473 and Zoom
14.01.2022 (10:00)Aiswarya Raj MunappyMaturity Assessment of Data PipelinesConference Presentation PracticeTBAhttps://chalmers.zoom.us/j/4604877890
28.01.2022 (10:00)Peter SamoaaSource Code Representation for Deep Learning in Software Engineering The usage of deep learning (DL) approaches for software engineering has attracted much attention. However, in order to use DL, source code needs to be formatted to fit the the expected input form of DL models. This problem is known as source code representation. Source code can be represented via different approaches, most importantly the tree-based, text-based, and graph-based approaches. In this paper, we use a systematic literature review (SLR) to detailedly investigate the representation approaches adopted in 103 studies that use DL in the context of software engineering. We show that each way of representating source code can provide a different, yet orthogonal view of the same source code. Thus, different software engineering tasks might require different (combinations of) code representation approaches, depending on the nature and complexity of the task. Particularly, we show that it is crucial to define whether the DL approach requires lexical, syntactical, or semantic code information. Our analysis shows that a wide range of different representations and combinations of representations (hybrid representations) are used to solve a wide range of common software engineering problems. However, we also observe a lack of generalizability of the presented approaches to other tasks, and validation based on industrial datasets.https://chalmers.zoom.us/j/65808304299
25.02.2022 (10:00)Prof. Martin ShepperdThe Prevalence of Errors in Machine Learning ExperimentsPresentation by Guest ProfessorComputational experiments are the dominant paradigm to understand ML algorithms. Thus we build knowledge through sense-making of many results, but we need to be sure our experimental results are reliable. Our re-analysis of experiments found ~40% with inconsistent results and/or basic statistical errors. We all make errors, so (i) use open science to expose to scrutiny, (ii) try to avoid dichotomous inferencing methods and (iii) use meta-analysis with caution!https://chalmers.zoom.us/j/69940500821 (Password: 194230)
10.03.2022 (14.15)Alexander StotskyRecursive Versus Non-Recursive Richardson Algorithms: Systematic Overview, Unified Frameworks and Application to Electric Grid Power Quality Monitoring Sufficiently accurate, fast and computationally efficient solution of the system of linear equations is required in many estimation problems. Unified framework for recursive computationally efficient convergence accelerators and error models for a number of combinations of Richardson and Newton-Schulz iterations is developed. The algorithms were tested for detection of the sag and swell signatures in the voltage and current signals on real data in three-phase power system, see https://doi.org/10.1080/00051144.2022.2039989https://chalmers.zoom.us/j/3410968234, pass:443602
01.04.2022 (10:00)Susanne Stenberg and Håkan Burden (RISE)The EU AI Act and its relevance for CSE researchSeminarSo you think you know what an AI system is? Well think again - because the EU knows better. Or at least, their definition will be the law and if you cannot comply the fine can be the larger number between €30.000.000 and 6% of world-wide annual turnover. The regulation is referred to as the AI Act and defines AI as a technology, when AI poses a high risk and what is needed in terms of data quality in order to be compliant. And where to put your CE-mark after self-certification. There will probably be exempts for research and General Purpose AI but systems placed on the market under continuous development will be affected. The AI Act will thus have an impact on the collaborations you seek outside of academia. But also in the longer run if your theoretical contributions towards data and algorithms are used by someone else and will require your assistance to be compliant. The seminar will be organized by Susanne Stenberg and Håkan Burden. They are both at RISE and work together on multiple projects regarding policy development and digitalization together with public administration and pivate enterprises. Susanne, master of laws, has worked as a judge in District Courts and the Courts of Appeal for over a decade and then as Secretary of Inquiries on several legislative proposals, ranging from the rules of democracy to the legal basis of digitalization within the public sector. She joined RISE in 2020, adding legal perspectives and policy development into R&D projects. Håkan holds a PhD in Computer Science from the University of Gothenburg. He joined RISE in 2015, initially in projects related to developing systems supporting open innovation within the automotive business. He has since then shifted his focus towards how regulations support and hinder business development as the digital sector expands to new areas.Hybrid: Jupiter 473 and https://chalmers.zoom.us/j/69101011658 (Password: 842702)
08.04.2022 (10:00)Weixing ZhangXtext-GrammarOptimizerPrepare for conferenceUsing Xtext to automatically generate DSL grammars from existing ecore metamodels is a well-known and important technique for designing DSLs. When we get the DSL syntax in this way, it is often not user-friendly and difficult to use, so DSL developers often need to manually modify and adapt it to make it easier to use, but for example, if you want to remove all brackets in a large-scale DSL, this will be a laborious work, and there may be many such cases in the adaptation of a DSL. This paper designs a grammar optimization tool that can be used according to the user's configuration to modify and adapt the target Xtext grammar in batches to make it the grammar that people ultimately want, thereby improving the developer's efficiency.https://gu-se.zoom.us/j/9143028235
13.05.2022 (10:00)Chi ZhangPredicting Pedestrian Behavior in Urban Traffic Scenarios Using Deep Learning MethodsLicentiate Trial Talk Hybrid: Jupiter 473 and https://gu-se.zoom.us/j/69217409577?pwd=L3B3dndoM0NvUWg4OS82NDFhR3FpQT09 (Password: 048868)

Registered students in 21 / 22:

  • Hazem Samoaa
  • Aiswarya Raj Munappy
  • Weixing Zhang
  • Hamdy Ayas
  • Khan Mohammad Habibullah
  • Linda Erlenhov