Alexander Schill
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Alexander Schill has a total of 62 co-authors. Here are the top co-authors by number of shared publications:
- Daniel Schuster
Technische Universität Dresden, Germany
4 shared publications - Ilja Shmelkin
Technische Universität Dresden
4 shared publications - Thomas Springer
Institute of Systems Architecture Chair of Distributed and Networked Systems TU Dresden
4 shared publications - Iris Braun
Technische Universität Dresden, Germany
3 shared publications - Josef Spillner
Zürich University of Applied Sciences, Winterthur, Switzerland
3 shared publications - Alexander Döschl
HM Hochschule München University of Applied Sciences
2 shared publications
Latest Publications
- Open Access
- 20 August 2024
- Published by Springer Nature in Social Network Analysis and Mining
- Vol. 14 (1) , 1-19
Intelligent algorithm selection for efficient update predictions in social media feeds
Efficiently synchronizing data with social media feeds while minimizing unnecessary requests presents a significant challenge in various fields. This paper investigates prediction algorithms for determining the optimal update intervals for Facebook and Twitter (now X) feeds, focusing on metrics such as delay (the time between a post’s publication and its retrieval) and requests per post. Variations in update intervals result in different algorithms producing varying results, making the selection of the most suitable algorithm for each feed crucial yet time-intensive. To address this, we propose three strategies for algorithm selection: baseline (applying a single algorithm to all feeds), optimum (identifying the best algorithm for each individual feed), and classification (selecting algorithms through a classification process based on each feed’s unique update patterns and context). Our strategies leverage various prediction algorithms, including static and adaptive algorithms, and inhomogeneous Poisson processes. We evaluate these strategies using real-world data from Facebook and Twitter, thoroughly assessing their performance in terms of delay and request efficiency. The findings demonstrate that the strategy Optimum effectively identifies the best algorithms for each feed, ensuring the highest prediction quality, though at a considerable computational cost. On the other hand, the strategy Classification offers superior runtime performance required to select algorithms. This research highlights the trade-offs between delay and request efficiency and presents a comprehensive solution for optimizing update predictions in social media feeds.
- 20 August 2024
- Published by Springer Nature in Social Network Analysis and Mining
- Vol. 14 (1) , 1-19
Open AccessIntelligent algorithm selection for efficient update predictions in social media feeds
Efficiently synchronizing data with social media feeds while minimizing unnecessary requests presents a significant challenge in various fields. This paper investigates prediction algorithms for determining the optimal update intervals for Facebook and Twitter (now X) feeds, focusing on metrics such as delay (the time between a post’s publication and its retrieval) and requests per post. Variations in update intervals result in different algorithms producing varying results, making the selection of the most suitable algorithm for each feed crucial yet time-intensive. To address this, we propose three strategies for algorithm selection: baseline (applying a single algorithm to all feeds), optimum (identifying the best algorithm for each individual feed), and classification (selecting algorithms through a classification process based on each feed’s unique update patterns and context). Our strategies leverage various prediction algorithms, including static and adaptive algorithms, and inhomogeneous Poisson processes. We evaluate these strategies using real-world data from Facebook and Twitter, thoroughly assessing their performance in terms of delay and request efficiency. The findings demonstrate that the strategy Optimum effectively identifies the best algorithms for each feed, ensuring the highest prediction quality, though at a considerable computational cost. On the other hand, the strategy Classification offers superior runtime performance required to select algorithms. This research highlights the trade-offs between delay and request efficiency and presents a comprehensive solution for optimizing update predictions in social media feeds.
- Open Access
- 6 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 340-344
Towards an adaptive algorithms selection on predicting the update interval for social media feeds
Efficiently synchronizing data with external sources, such as social media feeds, while minimizing well-timed requests is a challenge in various domains. This research investigates prediction algorithms for determining appropriate update intervals for Facebook and Twitter feeds, considering metrics such as the delay (time between a post's publication and retrieval) and requests per post. Due to variations in update intervals, different algorithms yield diverse results. Selecting the most suitable algorithm for each feed is a time-consuming but crucial task for achieving optimal resource usage. We propose three strategies for algorithm selection: baseline (using a single algorithm per feed), optimum (calculating the best algorithm for each feed), and classification (identifying algorithms through classification). Real-world data from Facebook and Twitter are used to evaluate the strategies, comprehensively assessing their strengths and weaknesses. Findings demonstrate that the strategy optimum identifies the best algorithms, while the strategy classification selects fairly good algorithms at significantly reduced computational effort.
- 6 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 340-344
Open AccessTowards an adaptive algorithms selection on predicting the update interval for social media feeds
Efficiently synchronizing data with external sources, such as social media feeds, while minimizing well-timed requests is a challenge in various domains. This research investigates prediction algorithms for determining appropriate update intervals for Facebook and Twitter feeds, considering metrics such as the delay (time between a post's publication and retrieval) and requests per post. Due to variations in update intervals, different algorithms yield diverse results. Selecting the most suitable algorithm for each feed is a time-consuming but crucial task for achieving optimal resource usage. We propose three strategies for algorithm selection: baseline (using a single algorithm per feed), optimum (calculating the best algorithm for each feed), and classification (identifying algorithms through classification). Real-world data from Facebook and Twitter are used to evaluate the strategies, comprehensively assessing their strengths and weaknesses. Findings demonstrate that the strategy optimum identifies the best algorithms, while the strategy classification selects fairly good algorithms at significantly reduced computational effort.
- 3 February 2023
- Published by Springer Nature
- p. 21-48
A Framework for Context-Sensitive Control-Loops with Roles
- 3 February 2023
- Published by Springer Nature
- p. 21-48
A Framework for Context-Sensitive Control-Loops with Roles
- 18 November 2022
- Published by Springer Nature
- p. 35-60
Adaptation Consistency of Distributed Role-Oriented Applications Based on the Actor Model of Computation
- 18 November 2022
- Published by Springer Nature
- p. 35-60
Adaptation Consistency of Distributed Role-Oriented Applications Based on the Actor Model of Computation
- 18 November 2022
- Published by Springer Nature
- p. 3-34
Towards Role-Based Context-Aware Monitoring Systems
- 18 November 2022
- Published by Springer Nature
- p. 3-34
Towards Role-Based Context-Aware Monitoring Systems
- 23 October 2022
- Published by Springer Nature
- p. 245-260
Towards Implementing Collaborative Learning in Remote Teaching Scenarios
- 23 October 2022
- Published by Springer Nature
- p. 245-260
Towards Implementing Collaborative Learning in Remote Teaching Scenarios
- 18 June 2020
- Published by Springer Nature
- p. 64-76
Peer-to-Peer Transfers for Crowd Monitoring - A Reality Check
- 18 June 2020
- Published by Springer Nature
- p. 64-76
Peer-to-Peer Transfers for Crowd Monitoring - A Reality Check
- 1 September 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
Towards a Development Methodology for Adaptable Collaborative Audience Response Systems
The use of Audience Response Systems (ARSs) for tech-enhanced learning scenarios has proven to address issues occurring within higher education, e.g. the missing interaction between the lecturer and the students. Since the majority of these systems relies on a single supported didactic concept and therefore has a limited set of provided functions, ARSs are currently restricted to support classic content-based as well as enquiry-based learning. The support of more advanced didactic concepts in order to investigate studio-based learning is currently not possible due to the lacking collaborative and cooperative functionality. This paper presents a unified (meta-)model which is able to express various scenarios, targeting the holistic support of content-based, enquiry-based and studio-based learning. The created model is evaluated within a user study to reason about the applicability of its underlying concept as well as the defined function blocks. In addition, this paper purposes ideas for a future graphical editor, which will support the modeling process, and provides concrete details for a possible implementation of a system on top of the (meta-)model.
- 1 September 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
Towards a Development Methodology for Adaptable Collaborative Audience Response Systems
The use of Audience Response Systems (ARSs) for tech-enhanced learning scenarios has proven to address issues occurring within higher education, e.g. the missing interaction between the lecturer and the students. Since the majority of these systems relies on a single supported didactic concept and therefore has a limited set of provided functions, ARSs are currently restricted to support classic content-based as well as enquiry-based learning. The support of more advanced didactic concepts in order to investigate studio-based learning is currently not possible due to the lacking collaborative and cooperative functionality. This paper presents a unified (meta-)model which is able to express various scenarios, targeting the holistic support of content-based, enquiry-based and studio-based learning. The created model is evaluated within a user study to reason about the applicability of its underlying concept as well as the defined function blocks. In addition, this paper purposes ideas for a future graphical editor, which will support the modeling process, and provides concrete details for a possible implementation of a system on top of the (meta-)model.
- Open Access
- 27 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 29 (11) , 2478-2488
Extending the Cutting Stock Problem for Consolidating Services with Stochastic Workloads
Data centres and similar sever clusters consume a large amount of energy. However, not all consumed energy produces a useful work. Servers consume a disproportional amount of energy when they are idle, underutilised, or overloaded. The effect of these conditions can be minimised by attempting to balance the demand for and the supply of resources through a careful prediction of future workloads and their efficient consolidation. In this paper we extend the cutting stock problem for consolidating workloads having stochastic characteristics. Hence, we employ the aggregate probability density function of co-located and simultaneously executing services to establish valid patterns. A valid pattern is one yielding an overall resource utilisation below a set threshold.We tested the scope and usefulness of our approach on a 16-core server with 29 different benchmarks. The workloads of these benchmarks have been generated based on the CPU utilisation traces of 100 real-world virtual machines which we obtained from a Google data centre hosting more than 32 000 virtual machines. Altogether, we considered 540 different consolidation scenarios during our experiment. We compared the performance of our approach — system overload probability, job completion time, and energy consumption — with four existing/proposed scheduling strategies. Thus, in each category, our approach incurred a modest penalty with respect to the best performing approach in that category, but overall resulted in a remarkable performance clearly demonstrating its capacity to achieve the best trade-off between resource consumption and performance.
- 27 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 29 (11) , 2478-2488
Open AccessExtending the Cutting Stock Problem for Consolidating Services with Stochastic Workloads
Data centres and similar sever clusters consume a large amount of energy. However, not all consumed energy produces a useful work. Servers consume a disproportional amount of energy when they are idle, underutilised, or overloaded. The effect of these conditions can be minimised by attempting to balance the demand for and the supply of resources through a careful prediction of future workloads and their efficient consolidation. In this paper we extend the cutting stock problem for consolidating workloads having stochastic characteristics. Hence, we employ the aggregate probability density function of co-located and simultaneously executing services to establish valid patterns. A valid pattern is one yielding an overall resource utilisation below a set threshold.We tested the scope and usefulness of our approach on a 16-core server with 29 different benchmarks. The workloads of these benchmarks have been generated based on the CPU utilisation traces of 100 real-world virtual machines which we obtained from a Google data centre hosting more than 32 000 virtual machines. Altogether, we considered 540 different consolidation scenarios during our experiment. We compared the performance of our approach — system overload probability, job completion time, and energy consumption — with four existing/proposed scheduling strategies. Thus, in each category, our approach incurred a modest penalty with respect to the best performing approach in that category, but overall resulted in a remarkable performance clearly demonstrating its capacity to achieve the best trade-off between resource consumption and performance.
- Open Access
- 1 January 2018
- Published by INSTICC
- p. 302-307
Multi-level De-anonymisation for Initially Anonymous Discussion Systems in a Self-regulated Learning Environment
- 1 January 2018
- Published by INSTICC
- p. 302-307
Open AccessMulti-level De-anonymisation for Initially Anonymous Discussion Systems in a Self-regulated Learning Environment