004 Datenverarbeitung; Informatik
Refine
Year of publication
- 2021 (78) (remove)
Document Type
- Article (58)
- Doctoral Thesis (7)
- Monograph/Edited Volume (6)
- Conference Proceeding (5)
- Postprint (2)
Is part of the Bibliography
- yes (78) (remove)
Keywords
- blockchain (2)
- business process management (2)
- business processes (2)
- computer vision (2)
- cyber-physical systems (2)
- deferred choice (2)
- formal semantics (2)
- memory (2)
- oracles (2)
- perception of robots (2)
Institute
- Hasso-Plattner-Institut für Digital Engineering GmbH (23)
- Hasso-Plattner-Institut für Digital Engineering gGmbH (19)
- Institut für Informatik und Computational Science (11)
- Fachgruppe Betriebswirtschaftslehre (7)
- Bürgerliches Recht (4)
- Department Linguistik (3)
- Extern (3)
- Institut für Biochemie und Biologie (3)
- Department Erziehungswissenschaft (2)
- Institut für Physik und Astronomie (2)
- Wirtschaftswissenschaften (2)
- Department Psychologie (1)
- Department Sport- und Gesundheitswissenschaften (1)
- Institut für Mathematik (1)
- Institut für Umweltwissenschaften und Geographie (1)
- Potsdam Transfer - Zentrum für Gründung, Innovation, Wissens- und Technologietransfer (1)
- Öffentliches Recht (1)
Cyber warfare is a timely and relevant issue and one of the most controversial in international humanitarian law (IHL). The aim of IHL is to set rules and limits in terms of means and methods of warfare. In this context, a key question arises: Has digital warfare rules or limits, and if so, how are these applicable? Traditional principles, developed over a long period, are facing a new dimension of challenges due to the rise of cyber warfare. This paper argues that to overcome this new issue, it is critical that new humanity-oriented approaches is developed with regard to cyber warfare. The challenge is to establish a legal regime for cyber-attacks, successfully addressing human rights norms and standards. While clarifying this from a legal perspective, the authors can redesign the sensitive equilibrium between humanity and military necessity, weighing the humanitarian aims of IHL and the protection of civilians-in combination with international human rights law and other relevant legal regimes-in a different manner than before.
Argument mining on twitter
(2021)
In the last decade, the field of argument mining has grown notably. However, only relatively few studies have investigated argumentation in social media and specifically on Twitter. Here, we provide the, to our knowledge, first critical in-depth survey of the state of the art in tweet-based argument mining. We discuss approaches to modelling the structure of arguments in the context of tweet corpus annotation, and we review current progress in the task of detecting argument components and their relations in tweets. We also survey the intersection of argument mining and stance detection, before we conclude with an outlook.
We systematically explore the effect of calibration data length on the performance of a conceptual hydrological model, GR4H, in comparison to two Artificial Neural Network (ANN) architectures: Long Short-Term Memory Networks (LSTM) and Gated Recurrent Units (GRU), which have just recently been introduced to the field of hydrology. We implemented a case study for six river basins across the contiguous United States, with 25 years of meteorological and discharge data. Nine years were reserved for independent validation; two years were used as a warm-up period, one year for each of the calibration and validation periods, respectively; from the remaining 14 years, we sampled increasing amounts of data for model calibration, and found pronounced differences in model performance. While GR4H required less data to converge, LSTM and GRU caught up at a remarkable rate, considering their number of parameters. Also, LSTM and GRU exhibited the higher calibration instability in comparison to GR4H. These findings confirm the potential of modern deep-learning architectures in rainfall runoff modelling, but also highlight the noticeable differences between them in regard to the effect of calibration data length.
ATIB
(2021)
Identity management is a principle component of securing online services. In the advancement of traditional identity management patterns, the identity provider remained a Trusted Third Party (TTP). The service provider and the user need to trust a particular identity provider for correct attributes amongst other demands. This paradigm changed with the invention of blockchain-based Self-Sovereign Identity (SSI) solutions that primarily focus on the users. SSI reduces the functional scope of the identity provider to an attribute provider while enabling attribute aggregation. Besides that, the development of new protocols, disregarding established protocols and a significantly fragmented landscape of SSI solutions pose considerable challenges for an adoption by service providers. We propose an Attribute Trust-enhancing Identity Broker (ATIB) to leverage the potential of SSI for trust-enhancing attribute aggregation. Furthermore, ATIB abstracts from a dedicated SSI solution and offers standard protocols. Therefore, it facilitates the adoption by service providers. Despite the brokered integration approach, we show that ATIB provides a high security posture. Additionally, ATIB does not compromise the ten foundational SSI principles for the users.
Cyber-physical systems often encompass complex concurrent behavior with timing constraints and probabilistic failures on demand. The analysis whether such systems with probabilistic timed behavior adhere to a given specification is essential. When the states of the system can be represented by graphs, the rule-based formalism of Probabilistic Timed Graph Transformation Systems (PTGTSs) can be used to suitably capture structure dynamics as well as probabilistic and timed behavior of the system. The model checking support for PTGTSs w.r.t. properties specified using Probabilistic Timed Computation Tree Logic (PTCTL) has been already presented. Moreover, for timed graph-based runtime monitoring, Metric Temporal Graph Logic (MTGL) has been developed for stating metric temporal properties on identified subgraphs and their structural changes over time. In this paper, we (a) extend MTGL to the Probabilistic Metric Temporal Graph Logic (PMTGL) by allowing for the specification of probabilistic properties, (b) adapt our MTGL satisfaction checking approach to PTGTSs, and (c) combine the approaches for PTCTL model checking and MTGL satisfaction checking to obtain a Bounded Model Checking (BMC) approach for PMTGL. In our evaluation, we apply an implementation of our BMC approach in AutoGraph to a running example.
Gene expression data provide the expression levels of tens of thousands of genes from several hundred samples. These data are analyzed to detect biomarkers that can be of prognostic or diagnostic use. Traditionally, biomarker detection for gene expression data is the task of gene selection. The vast number of genes is reduced to a few relevant ones that achieve the best performance for the respective use case. Traditional approaches select genes based on their statistical significance in the data set. This results in issues of robustness, redundancy and true biological relevance of the selected genes. Integrative analyses typically address these shortcomings by integrating multiple data artifacts from the same objects, e.g. gene expression and methylation data. When only gene expression data are available, integrative analyses instead use curated information on biological processes from public knowledge bases. With knowledge bases providing an ever-increasing amount of curated biological knowledge, such prior knowledge approaches become more powerful. This paper provides a thorough overview on the status quo of biomarker detection on gene expression data with prior biological knowledge. We discuss current shortcomings of traditional approaches, review recent external knowledge bases, provide a classification and qualitative comparison of existing prior knowledge approaches and discuss open challenges for this kind of gene selection.
Many important graph-theoretic notions can be encoded as counting graph homomorphism problems, such as partition functions in statistical physics, in particular independent sets and colourings. In this article, we study the complexity of #(p) HOMSTOH, the problem of counting graph homomorphisms from an input graph to a graph H modulo a prime number p. Dyer and Greenhill proved a dichotomy stating that the tractability of non-modular counting graph homomorphisms depends on the structure of the target graph. Many intractable cases in non-modular counting become tractable in modular counting due to the common phenomenon of cancellation. In subsequent studies on counting modulo 2, however, the influence of the structure of H on the tractability was shown to persist, which yields similar dichotomies. <br /> Our main result states that for every tree H and every prime p the problem #pHOMSTOH is either polynomial time computable or #P-p-complete. This relates to the conjecture of Faben and Jerrum stating that this dichotomy holds for every graph H when counting modulo 2. In contrast to previous results on modular counting, the tractable cases of #pHOMSTOH are essentially the same for all values of the modulo when H is a tree. To prove this result, we study the structural properties of a homomorphism. As an important interim result, our study yields a dichotomy for the problem of counting weighted independent sets in a bipartite graph modulo some prime p. These results are the first suggesting that such dichotomies hold not only for the modulo 2 case but also for the modular counting functions of all primes p.
Spreadsheets are among the most commonly used file formats for data management, distribution, and analysis. Their widespread employment makes it easy to gather large collections of data, but their flexible canvas-based structure makes automated analysis difficult without heavy preparation. One of the common problems that practitioners face is the presence of multiple, independent regions in a single spreadsheet, possibly separated by repeated empty cells. We define such files as "multiregion" files. In collections of various spreadsheets, we can observe that some share the same layout. We present the Mondrian approach to automatically identify layout templates across multiple files and systematically extract the corresponding regions. Our approach is composed of three phases: first, each file is rendered as an image and inspected for elements that could form regions; then, using a clustering algorithm, the identified elements are grouped to form regions; finally, every file layout is represented as a graph and compared with others to find layout templates. We compare our method to state-of-the-art table recognition algorithms on two corpora of real-world enterprise spreadsheets. Our approach shows the best performances in detecting reliable region boundaries within each file and can correctly identify recurring layouts across files.
The automated detection of sequential anomalies in time series is an essential task for many applications, such as the monitoring of technical systems, fraud detection in high-frequency trading, or the early detection of disease symptoms. All these applications require the detection to find all sequential anomalies possibly fast on potentially very large time series. In other words, the detection needs to be effective, efficient and scalable w.r.t. the input size. Series2Graph is an effective solution based on graph embeddings that are robust against re-occurring anomalies and can discover sequential anomalies of arbitrary length and works without training data. Yet, Series2Graph is no t scalable due to its single-threaded approach; it cannot, in particular, process arbitrarily large sequences due to the memory constraints of a single machine. In this paper, we propose our distributed anomaly detection system, short DADS, which is an efficient and scalable adaptation of Series2Graph. Based on the actor programming model, DADS distributes the input time sequence, intermediate state and the computation to all processors of a cluster in a way that minimizes communication costs and synchronization barriers. Our evaluation shows that DADS is orders of magnitude faster than S2G, scales almost linearly with the number of processors in the cluster and can process much larger input sequences due to its scale-out property.
The integration of multiple data sources is a common problem in a large variety of applications. Traditionally, handcrafted similarity measures are used to discover, merge, and integrate multiple representations of the same entity-duplicates-into a large homogeneous collection of data. Often, these similarity measures do not cope well with the heterogeneity of the underlying dataset. In addition, domain experts are needed to manually design and configure such measures, which is both time-consuming and requires extensive domain expertise. <br /> We propose a deep Siamese neural network, capable of learning a similarity measure that is tailored to the characteristics of a particular dataset. With the properties of deep learning methods, we are able to eliminate the manual feature engineering process and thus considerably reduce the effort required for model construction. In addition, we show that it is possible to transfer knowledge acquired during the deduplication of one dataset to another, and thus significantly reduce the amount of data required to train a similarity measure. We evaluated our method on multiple datasets and compare our approach to state-of-the-art deduplication methods. Our approach outperforms competitors by up to +26 percent F-measure, depending on task and dataset. In addition, we show that knowledge transfer is not only feasible, but in our experiments led to an improvement in F-measure of up to +4.7 percent.
Viper
(2021)
Key-value stores (KVSs) have found wide application in modern software systems. For persistence, their data resides in slow secondary storage, which requires KVSs to employ various techniques to increase their read and write performance from and to the underlying medium. Emerging persistent memory (PMem) technologies offer data persistence at close-to-DRAM speed, making them a promising alternative to classical disk-based storage. However, simply drop-in replacing existing storage with PMem does not yield good results, as block-based access behaves differently in PMem than on disk and ignores PMem's byte addressability, layout, and unique performance characteristics. In this paper, we propose three PMem-specific access patterns and implement them in a hybrid PMem-DRAM KVS called Viper. We employ a DRAM-based hash index and a PMem-aware storage layout to utilize the random-write speed of DRAM and efficient sequential-write performance PMem. Our evaluation shows that Viper significantly outperforms existing KVSs for core KVS operations while providing full data persistence. Moreover, Viper outperforms existing PMem-only, hybrid, and disk-based KVSs by 4-18x for write workloads, while matching or surpassing their get performance.
Which event happened first?
(2021)
First come, first served: Critical choices between alternative actions are often made based on events external to an organization, and reacting promptly to their occurrence can be a major advantage over the competition. In Business Process Management (BPM), such deferred choices can be expressed in process models, and they are an important aspect of process engines. Blockchain-based process execution approaches are no exception to this, but are severely limited by the inherent properties of the platform: The isolated environment prevents direct access to external entities and data, and the non-continual runtime based entirely on atomic transactions impedes the monitoring and detection of events. In this paper we provide an in-depth examination of the semantics of deferred choice, and transfer them to environments such as the blockchain. We introduce and compare several oracle architectures able to satisfy certain requirements, and show that they can be implemented using state-of-the-art blockchain technology.
First come, first served: Critical choices between alternative actions are often made based on events external to an organization, and reacting promptly to their occurrence can be a major advantage over the competition. In Business Process Management (BPM), such deferred choices can be expressed in process models, and they are an important aspect of process engines. Blockchain-based process execution approaches are no exception to this, but are severely limited by the inherent properties of the platform: The isolated environment prevents direct access to external entities and data, and the non-continual runtime based entirely on atomic transactions impedes the monitoring and detection of events. In this paper we provide an in-depth examination of the semantics of deferred choice, and transfer them to environments such as the blockchain. We introduce and compare several oracle architectures able to satisfy certain requirements, and show that they can be implemented using state-of-the-art blockchain technology.
Background:
Research into the application of virtual reality technology in the health care sector has rapidly increased, resulting in a large body of research that is difficult to keep up with.
Objective:
We will provide an overview of the annual publication numbers in this field and the most productive and influential countries, journals, and authors, as well as the most used, most co-occurring, and most recent keywords.
Methods:
Based on a data set of 356 publications and 20,363 citations derived from Web of Science, we conducted a bibliometric analysis using BibExcel, HistCite, and VOSviewer.
Results:
The strongest growth in publications occurred in 2020, accounting for 29.49% of all publications so far. The most productive countries are the United States, the United Kingdom, and Spain; the most influential countries are the United States, Canada, and the United Kingdom. The most productive journals are the Journal of Medical Internet Research (JMIR), JMIR Serious Games, and the Games for Health Journal; the most influential journals are Patient Education and Counselling, Medical Education, and Quality of Life Research. The most productive authors are Riva, del Piccolo, and Schwebel; the most influential authors are Finset, del Piccolo, and Eide. The most frequently occurring keywords other than “virtual” and “reality” are “training,” “trial,” and “patients.” The most relevant research themes are communication, education, and novel treatments; the most recent research trends are fitness and exergames.
Conclusions:
The analysis shows that the field has left its infant state and its specialization is advancing, with a clear focus on patient usability.
Comprior
(2021)
Background
Reproducible benchmarking is important for assessing the effectiveness of novel feature selection approaches applied on gene expression data, especially for prior knowledge approaches that incorporate biological information from online knowledge bases. However, no full-fledged benchmarking system exists that is extensible, provides built-in feature selection approaches, and a comprehensive result assessment encompassing classification performance, robustness, and biological relevance. Moreover, the particular needs of prior knowledge feature selection approaches, i.e. uniform access to knowledge bases, are not addressed. As a consequence, prior knowledge approaches are not evaluated amongst each other, leaving open questions regarding their effectiveness.
Results
We present the Comprior benchmark tool, which facilitates the rapid development and effortless benchmarking of feature selection approaches, with a special focus on prior knowledge approaches. Comprior is extensible by custom approaches, offers built-in standard feature selection approaches, enables uniform access to multiple knowledge bases, and provides a customizable evaluation infrastructure to compare multiple feature selection approaches regarding their classification performance, robustness, runtime, and biological relevance.
Conclusion
Comprior allows reproducible benchmarking especially of prior knowledge approaches, which facilitates their applicability and for the first time enables a comprehensive assessment of their effectiveness
Generative adversarial networks (GANs) have been broadly applied to a wide range of application domains since their proposal. In this thesis, we propose several methods that aim to tackle different existing problems in GANs. Particularly, even though GANs are generally able to generate high-quality samples, the diversity of the generated set is often sub-optimal. Moreover, the common increase of the number of models in the original GANs framework, as well as their architectural sizes, introduces additional costs. Additionally, even though challenging, the proper evaluation of a generated set is an important direction to ultimately improve the generation process in GANs. We start by introducing two diversification methods that extend the original GANs framework to multiple adversaries to stimulate sample diversity in a generated set. Then, we introduce a new post-training compression method based on Monte Carlo methods and importance sampling to quantize and prune the weights and activations of pre-trained neural networks without any additional training. The previous method may be used to reduce the memory and computational costs introduced by increasing the number of models in the original GANs framework. Moreover, we use a similar procedure to quantize and prune gradients during training, which also reduces the communication costs between different workers in a distributed training setting. We introduce several topology-based evaluation methods to assess data generation in different settings, namely image generation and language generation. Our methods retrieve both single-valued and double-valued metrics, which, given a real set, may be used to broadly assess a generated set or separately evaluate sample quality and sample diversity, respectively. Moreover, two of our metrics use locality-sensitive hashing to accurately assess the generated sets of highly compressed GANs. The analysis of the compression effects in GANs paves the way for their efficient employment in real-world applications. Given their general applicability, the methods proposed in this thesis may be extended beyond the context of GANs. Hence, they may be generally applied to enhance existing neural networks and, in particular, generative frameworks.
3D point clouds are a universal and discrete digital representation of three-dimensional objects and environments. For geospatial applications, 3D point clouds have become a fundamental type of raw data acquired and generated using various methods and techniques. In particular, 3D point clouds serve as raw data for creating digital twins of the built environment.
This thesis concentrates on the research and development of concepts, methods, and techniques for preprocessing, semantically enriching, analyzing, and visualizing 3D point clouds for applications around transport infrastructure. It introduces a collection of preprocessing techniques that aim to harmonize raw 3D point cloud data, such as point density reduction and scan profile detection. Metrics such as, e.g., local density, verticality, and planarity are calculated for later use. One of the key contributions tackles the problem of analyzing and deriving semantic information in 3D point clouds. Three different approaches are investigated: a geometric analysis, a machine learning approach operating on synthetically generated 2D images, and a machine learning approach operating on 3D point clouds without intermediate representation.
In the first application case, 2D image classification is applied and evaluated for mobile mapping data focusing on road networks to derive road marking vector data. The second application case investigates how 3D point clouds can be merged with ground-penetrating radar data for a combined visualization and to automatically identify atypical areas in the data. For example, the approach detects pavement regions with developing potholes. The third application case explores the combination of a 3D environment based on 3D point clouds with panoramic imagery to improve visual representation and the detection of 3D objects such as traffic signs.
The presented methods were implemented and tested based on software frameworks for 3D point clouds and 3D visualization. In particular, modules for metric computation, classification procedures, and visualization techniques were integrated into a modular pipeline-based C++ research framework for geospatial data processing, extended by Python machine learning scripts. All visualization and analysis techniques scale to large real-world datasets such as road networks of entire cities or railroad networks.
The thesis shows that some use cases allow taking advantage of established image vision methods to analyze images rendered from mobile mapping data efficiently. The two presented semantic classification methods working directly on 3D point clouds are use case independent and show similar overall accuracy when compared to each other. While the geometry-based method requires less computation time, the machine learning-based method supports arbitrary semantic classes but requires training the network with ground truth data. Both methods can be used in combination to gradually build this ground truth with manual corrections via a respective annotation tool.
This thesis contributes results for IT system engineering of applications, systems, and services that require spatial digital twins of transport infrastructure such as road networks and railroad networks based on 3D point clouds as raw data. It demonstrates the feasibility of fully automated data flows that map captured 3D point clouds to semantically classified models. This provides a key component for seamlessly integrated spatial digital twins in IT solutions that require up-to-date, object-based, and semantically enriched information about the built environment.
Die stetige Weiterentwicklung von VR-Systemen bietet neue Möglichkeiten der Interaktion mit virtuellen Objekten im dreidimensionalen Raum, stellt Entwickelnde von VRAnwendungen aber auch vor neue Herausforderungen. Selektions- und Manipulationstechniken müssen unter Berücksichtigung des Anwendungsszenarios, der Zielgruppe und der zur Verfügung stehenden Ein- und Ausgabegeräte ausgewählt werden. Diese Arbeit leistet einen Beitrag dazu, die Auswahl von passenden Interaktionstechniken zu unterstützen. Hierfür wurde eine repräsentative Menge von Selektions- und Manipulationstechniken untersucht und, unter Berücksichtigung existierender Klassifikationssysteme, eine Taxonomie entwickelt, die die Analyse der Techniken hinsichtlich interaktionsrelevanter Eigenschaften ermöglicht. Auf Basis dieser Taxonomie wurden Techniken ausgewählt, die in einer explorativen Studie verglichen wurden, um Rückschlüsse auf die Dimensionen der Taxonomie zu ziehen und neue Indizien für Vor- und Nachteile der Techniken in spezifischen Anwendungsszenarien zu generieren. Die Ergebnisse der Arbeit münden in eine Webanwendung, die Entwickelnde von VR-Anwendungen gezielt dabei unterstützt, passende Selektions- und Manipulationstechniken für ein Anwendungsszenario auszuwählen, indem Techniken auf Basis der Taxonomie gefiltert und unter Verwendung der Resultate aus der Studie sortiert werden können.
The formal modeling and analysis is of crucial importance for software development processes following the model based approach. We present the formalism of Interval Probabilistic Timed Graph Transformation Systems (IPTGTSs) as a high-level modeling language. This language supports structure dynamics (based on graph transformation), timed behavior (based on clocks, guards, resets, and invariants as in Timed Automata (TA)), and interval probabilistic behavior (based on Discrete Interval Probability Distributions). That is, for the probabilistic behavior, the modeler using IPTGTSs does not need to provide precise probabilities, which are often impossible to obtain, but rather provides a probability range instead from which a precise probability is chosen nondeterministically. In fact, this feature on capturing probabilistic behavior distinguishes IPTGTSs from Probabilistic Timed Graph Transformation Systems (PTGTSs) presented earlier.
Following earlier work on Interval Probabilistic Timed Automata (IPTA) and PTGTSs, we also provide an analysis tool chain for IPTGTSs based on inter-formalism transformations. In particular, we provide in our tool AutoGraph a translation of IPTGTSs to IPTA and rely on a mapping of IPTA to Probabilistic Timed Automata (PTA) to allow for the usage of the Prism model checker. The tool Prism can then be used to analyze the resulting PTA w.r.t. probabilistic real-time queries asking for worst-case and best-case probabilities to reach a certain set of target states in a given amount of time.
Learning analytics at scale
(2021)
Digital technologies are paving the way for innovative educational approaches. The learning format of Massive Open Online Courses (MOOCs) provides a highly accessible path to lifelong learning while being more affordable and flexible than face-to-face courses. Thereby, thousands of learners can enroll in courses mostly without admission restrictions, but this also raises challenges. Individual supervision by teachers is barely feasible, and learning persistence and success depend on students' self-regulatory skills. Here, technology provides the means for support. The use of data for decision-making is already transforming many fields, whereas in education, it is still a young research discipline. Learning Analytics (LA) is defined as the measurement, collection, analysis, and reporting of data about learners and their learning contexts with the purpose of understanding and improving learning and learning environments. The vast amount of data that MOOCs produce on the learning behavior and success of thousands of students provides the opportunity to study human learning and develop approaches addressing the demands of learners and teachers.
The overall purpose of this dissertation is to investigate the implementation of LA at the scale of MOOCs and to explore how data-driven technology can support learning and teaching in this context. To this end, several research prototypes have been iteratively developed for the HPI MOOC Platform. Hence, they were tested and evaluated in an authentic real-world learning environment. Most of the results can be applied on a conceptual level to other MOOC platforms as well. The research contribution of this thesis thus provides practical insights beyond what is theoretically possible. In total, four system components were developed and extended:
(1) The Learning Analytics Architecture: A technical infrastructure to collect, process, and analyze event-driven learning data based on schema-agnostic pipelining in a service-oriented MOOC platform. (2) The Learning Analytics Dashboard for Learners: A tool for data-driven support of self-regulated learning, in particular to enable learners to evaluate and plan their learning activities, progress, and success by themselves. (3) Personalized Learning Objectives: A set of features to better connect learners' success to their personal intentions based on selected learning objectives to offer guidance and align the provided data-driven insights about their learning progress. (4) The Learning Analytics Dashboard for Teachers: A tool supporting teachers with data-driven insights to enable the monitoring of their courses with thousands of learners, identify potential issues, and take informed action.
For all aspects examined in this dissertation, related research is presented, development processes and implementation concepts are explained, and evaluations are conducted in case studies. Among other findings, the usage of the learner dashboard in combination with personalized learning objectives demonstrated improved certification rates of 11.62% to 12.63%. Furthermore, it was observed that the teacher dashboard is a key tool and an integral part for teaching in MOOCs. In addition to the results and contributions, general limitations of the work are discussed—which altogether provide a solid foundation for practical implications and future research.
In recent years, computer vision algorithms based on machine learning have seen rapid development. In the past, research mostly focused on solving computer vision problems such as image classification or object detection on images displaying natural scenes. Nowadays other fields such as the field of cultural heritage, where an abundance of data is available, also get into the focus of research. In the line of current research endeavours, we collaborated with the Getty Research Institute which provided us with a challenging dataset, containing images of paintings and drawings. In this technical report, we present the results of the seminar "Deep Learning for Computer Vision". In this seminar, students of the Hasso Plattner Institute evaluated state-of-the-art approaches for image classification, object detection and image recognition on the dataset of the Getty Research Institute. The main challenge when applying modern computer vision methods to the available data is the availability of annotated training data, as the dataset provided by the Getty Research Institute does not contain a sufficient amount of annotated samples for the training of deep neural networks. However, throughout the report we show that it is possible to achieve satisfying to very good results, when using further publicly available datasets, such as the WikiArt dataset, for the training of machine learning models.
The noble way to substantiate decisions that affect many people is to ask these people for their opinions. For governments that run whole countries, this means asking all citizens for their views to consider their situations and needs.
Organizations such as Africa's Voices Foundation, who want to facilitate communication between decision-makers and citizens of a country, have difficulty mediating between these groups. To enable understanding, statements need to be summarized and visualized. Accomplishing these goals in a way that does justice to the citizens' voices and situations proves challenging. Standard charts do not help this cause as they fail to create empathy for the people behind their graphical abstractions. Furthermore, these charts do not create trust in the data they are representing as there is no way to see or navigate back to the underlying code and the original data. To fulfill these functions, visualizations would highly benefit from interactions to explore the displayed data, which standard charts often only limitedly provide.
To help improve the understanding of people's voices, we developed and categorized 80 ideas for new visualizations, new interactions, and better connections between different charts, which we present in this report. From those ideas, we implemented 10 prototypes and two systems that integrate different visualizations. We show that this integration allows consistent appearance and behavior of visualizations. The visualizations all share the same main concept: representing each individual with a single dot. To realize this idea, we discuss technologies that efficiently allow the rendering of a large number of these dots. With these visualizations, direct interactions with representations of individuals are achievable by clicking on them or by dragging a selection around them. This direct interaction is only possible with a bidirectional connection from the visualization to the data it displays. We discuss different strategies for bidirectional mappings and the trade-offs involved. Having unified behavior across visualizations enhances exploration. For our prototypes, that includes grouping, filtering, highlighting, and coloring of dots. Our prototyping work was enabled by the development environment Lively4. We explain which parts of Lively4 facilitated our prototyping process. Finally, we evaluate our approach to domain problems and our developed visualization concepts.
Our work provides inspiration and a starting point for visualization development in this domain. Our visualizations can improve communication between citizens and their government and motivate empathetic decisions. Our approach, combining low-level entities to create visualizations, provides value to an explorative and empathetic workflow. We show that the design space for visualizing this kind of data has a lot of potential and that it is possible to combine qualitative and quantitative approaches to data analysis.
TRIPOD
(2021)
Inertial measurement units (IMUs) enable easy to operate and low-cost data recording for gait analysis. When combined with treadmill walking, a large number of steps can be collected in a controlled environment without the need of a dedicated gait analysis laboratory. In order to evaluate existing and novel IMU-based gait analysis algorithms for treadmill walking, a reference dataset that includes IMU data as well as reliable ground truth measurements for multiple participants and walking speeds is needed. This article provides a reference dataset consisting of 15 healthy young adults who walked on a treadmill at three different speeds. Data were acquired using seven IMUs placed on the lower body, two different reference systems (Zebris FDMT-HQ and OptoGait), and two RGB cameras. Additionally, in order to validate an existing IMU-based gait analysis algorithm using the dataset, an adaptable modular data analysis pipeline was built. Our results show agreement between the pressure-sensitive Zebris and the photoelectric OptoGait system (r = 0.99), demonstrating the quality of our reference data. As a use case, the performance of an algorithm originally designed for overground walking was tested on treadmill data using the data pipeline. The accuracy of stride length and stride time estimations was comparable to that reported in other studies with overground data, indicating that the algorithm is equally applicable to treadmill data. The Python source code of the data pipeline is publicly available, and the dataset will be provided by the authors upon request, enabling future evaluations of IMU gait analysis algorithms without the need of recording new data.
Crochet is a popular handcraft all over the world. While other techniques such as knitting or weaving have received technical support over the years through machines, crochet is still a purely manual craft. Not just the act of crochet itself is manual but also the process of creating instructions for new crochet patterns, which is barely supported by domain specific digital solutions. This leads to unstructured and often also ambiguous and erroneous pattern instructions. In this report, we propose a concept to digitally represent crochet patterns. This format incorporates crochet techniques which allows domain specific support for crochet pattern designers during the pattern creation and instruction writing process. As contributions, we present a thorough domain analysis, the concept of a graph structure used as domain specific language to specify crochet patterns and a prototype of a projectional editor using the graph as representation format of patterns and a diagramming system to visualize them in 2D and 3D. By analyzing the domain, we learned about crochet techniques and pain points of designers in their pattern creation workflow. These insights are the basis on which we defined the pattern representation. In order to evaluate our concept, we built a prototype by which the feasibility of the concept is shown and we tested the software with professional crochet designers who approved of the concept.
In the last decades, there was a notable progress in solving the well-known Boolean satisfiability (Sat) problem, which can be witnessed by powerful Sat solvers. One of the reasons why these solvers are so fast are structural properties of instances that are utilized by the solver’s interna. This thesis deals with the well-studied structural property treewidth, which measures the closeness of an instance to being a tree. In fact, there are many problems parameterized by treewidth that are solvable in polynomial time in the instance size when parameterized by treewidth.
In this work, we study advanced treewidth-based methods and tools for problems in knowledge representation and reasoning (KR). Thereby, we provide means to establish precise runtime results (upper bounds) for canonical problems relevant to KR. Then, we present a new type of problem reduction, which we call decomposition-guided (DG) that
allows us to precisely monitor the treewidth when reducing from one problem to another problem. This new reduction type will be the basis for a long-open lower bound result for quantified Boolean formulas and allows us to design a new methodology for establishing runtime lower bounds for problems parameterized by treewidth.
Finally, despite these lower bounds, we provide an efficient implementation of algorithms that adhere to treewidth. Our approach finds suitable abstractions of instances, which are subsequently refined in a recursive fashion, and it uses Sat solvers for solving subproblems. It turns out that our resulting solver is quite competitive for two canonical counting problems related to Sat.
Discriminative Models for Biometric Identification using Micro- and Macro-Movements of the Eyes
(2021)
Human visual perception is an active process. Eye movements either alternate between fixations and saccades or follow a smooth pursuit movement in case of moving targets. Besides these macroscopic gaze patterns, the eyes perform involuntary micro-movements during fixations which are commonly categorized into micro-saccades, drift and tremor. Eye movements are frequently studied in cognitive psychology, because they reflect a complex interplay of perception, attention and oculomotor control.
A common insight of psychological research is that macro-movements are highly individual. Inspired by this finding, there has been a considerable amount of prior research on oculomotoric biometric identification. However, the accuracy of known approaches is too low and the time needed for identification is too long for any practical application. This thesis explores discriminative models for the task of biometric identification.
Discriminative models optimize a quality measure of the predictions and are usually superior to generative approaches in discriminative tasks. However, using discriminative models requires to select a suitable form of data representation for sequential eye gaze data; i.e., by engineering features or constructing a sequence kernel and the performance of the classification model strongly depends on the data representation. We study two fundamentally different ways of representing eye gaze within a discriminative framework. In the first part of this thesis, we explore the integration of data and psychological background knowledge in the form of generative models to construct representations. To this end, we first develop generative statistical models of gaze behavior during reading and scene viewing that account for viewer-specific distributional properties of gaze patterns. In a second step, we develop a discriminative identification model by deriving Fisher kernel functions from these and several baseline models. We find that an SVM with Fisher kernel is able to reliably identify users based on their eye gaze during reading and scene viewing. However, since the generative models are constrained to use low-frequency macro-movements, they discard a significant amount of information contained in the raw eye tracking signal at a high cost: identification requires about one minute of input recording, which makes it inapplicable for real world biometric systems. In the second part of this thesis, we study a purely data-driven modeling approach. Here, we aim at automatically discovering the individual pattern hidden in the raw eye tracking signal. To this end, we develop a deep convolutional neural network DeepEyedentification that processes yaw and pitch gaze velocities and learns a representation end-to-end. Compared to prior work, this model increases the identification accuracy by one order of magnitude and the time to identification decreases to only seconds. The DeepEyedentificationLive model further improves upon the identification performance by processing binocular input and it also detects presentation-attacks.
We find that by learning a representation, the performance of oculomotoric identification and presentation-attack detection can be driven close to practical relevance for biometric applications. Eye tracking devices with high sampling frequency and precision are expensive and the applicability of eye movement as a biometric feature heavily depends on cost of recording devices.
In the last part of this thesis, we therefore study the requirements on data quality by evaluating the performance of the DeepEyedentificationLive network under reduced spatial and temporal resolution. We find that the method still attains a high identification accuracy at a temporal resolution of only 250 Hz and a precision of 0.03 degrees. Reducing both does not have an additive deteriorating effect.
EMOOCs 2021
(2021)
From June 22 to June 24, 2021, Hasso Plattner Institute, Potsdam, hosted the seventh European MOOC Stakeholder Summit (EMOOCs 2021) together with the eighth ACM Learning@Scale Conference.
Due to the COVID-19 situation, the conference was held fully online.
The boost in digital education worldwide as a result of the pandemic was also one of the main topics of this year’s EMOOCs. All institutions of learning have been forced to transform and redesign their educational methods, moving from traditional models to hybrid or completely online models at scale. The learnings, derived from practical experience and research, have been explored in EMOOCs 2021 in six tracks and additional workshops, covering various aspects of this field. In this publication, we present papers from the conference’s Experience Track, the Policy Track, the Business Track, the International Track, and the Workshops.
Compound values are not universally supported in virtual machine (VM)-based programming systems and languages. However, providing data structures with value characteristics can be beneficial. On one hand, programming systems and languages can adequately represent physical quantities with compound values and avoid inconsistencies, for example, in representation of large numbers. On the other hand, just-in-time (JIT) compilers, which are often found in VMs, can rely on the fact that compound values are immutable, which is an important property in optimizing programs. Considering this, compound values have an optimization potential that can be put to use by implementing them in VMs in a way that is efficient in memory usage and execution time. Yet, optimized compound values in VMs face certain challenges: to maintain consistency, it should not be observable by the program whether compound values are represented in an optimized way by a VM; an optimization should take into account, that the usage of compound values can exhibit certain patterns at run-time; and that necessary value-incompatible properties due to implementation restrictions should be reduced.
We propose a technique to detect and compress common patterns of compound value usage at run-time to improve memory usage and execution speed. Our approach identifies patterns of frequent compound value references and introduces abbreviated forms for them. Thus, it is possible to store multiple inter-referenced compound values in an inlined memory representation, reducing the overhead of metadata and object references. We extend our approach by a notion of limited mutability, using cells that act as barriers for our approach and provide a location for shared, mutable access with the possibility of type specialization. We devise an extension to our approach that allows us to express automatic unboxing of boxed primitive data types in terms of our initial technique. We show that our approach is versatile enough to express another optimization technique that relies on values, such as Booleans, that are unique throughout a programming system. Furthermore, we demonstrate how to re-use learned usage patterns and optimizations across program runs, thus reducing the performance impact of pattern recognition.
We show in a best-case prototype that the implementation of our approach is feasible and can also be applied to general purpose programming systems, namely implementations of the Racket language and Squeak/Smalltalk. In several micro-benchmarks, we found that our approach can effectively reduce memory consumption and improve execution speed.