1,721,114 research outputs found
Envisioning Tool Support for Designing Privacy-Aware Internet of Thing Applications
Get PDFThe design and development process for Internet of Things (IoT) applications is more complicated than for desktop, mobile, or web applications. IoT applications require both software and hardware to work together across multiple different types of nodes (e.g., microcontrollers, system-on-chips, mobile phones, miniaturized single-board computers, and cloud platforms) with different capabilities under different conditions. IoT applications typically collect and analyze personal data that can be used to derive sensitive information about individuals. Without proper privacy protections in place, IoT applications could lead to serious privacy violations. Thus far, privacy concerns have not been explicitly considered in software engineering processes when designing and developing IoT applications, partly due to a lack of tools, technologies, and guidance. This paper presents a research vision that argues the importance of developing a privacy-aware IoT application design tool to address the challenges mentioned above. This tool should not only transform IoT application designs into privacy-aware application designs, but also validate and verify them. First, we outline how this proposed tool should work in practice and its core functionalities. Then, we identify research challenges and potential directions for developing the proposed tool. We anticipate that this proposed tool will save many engineering hours which engineers would otherwise need to spend on developing privacy expertise and applying it. We also highlight the usefulness of this tool toward privacy education and privacy compliance
Crowdsourced Peer Learning Activity for Internet of Things Education: A Case Study
No full textComputing devices such as laptops, tablets and mobile phones have become part of our daily lives. End users increasingly know more and more information about these devices. Further, more technically savvy end users know how such devices are being built and know how to choose one over the others. However, we cannot say the same about the Internet of Things (IoT) products. Due to its infancy nature of the marketplace, end users have very little idea about IoT products. To address this issue, we developed a method, a crowdsourced peer learning activity, supported by an online platform (OLYMPUS) to enable a group of learners to learn IoT products space better. We conducted two different user studies to validate that our tool enables better IoT education. Our method guide learners to think more deeply about IoT products and their design decisions. The learning platform we developed is open source and available for the community
FINE-GRAINED TAINT TRACING IN CRYPTOCURRENCIES: A SUB-RANGE-BASED UTXO MODEL WITH VERKLE TREE PROOFS
No full textCryptocurrency taint analysis is critical for tracking illicit funds, but existing methods either over-taint legitimate coins or struggle with scalability. This thesis proposes a finegrained taint tracing model for UTXOs that treats each coin as a collection of indivisible sub-units (satoshis) with unique IDs. We design a sub-range-based UTXO model with the support of Verkle tree commitments to efficiently encode and prove ownership of these sub-units and to validate the correct computation. Our approach allows a user to spend arbitrary portions of a UTXO while providing succinct cryptographic proofs: a membership proof to showeach spent sub-unit existed in theUTXO, a delta proof to prove the UTXO's commitment updates correctly after removing spent parts, and a combined Vroot proof to verify multiple inputs merge consistently into newUTXO.We validate the model's correctness using Kate-Zaverucha-Goldberg (KZG) polynomial commitments ensuring that the sum of spent and remaining parts matches the original commitment with minimal verification overhead. In our design, no on-chain enumeration of sub-units is required , only short commitments and proofs. Key findings demonstrate that finegrained taint can be traced through complex transactions (including coin mixes) without false positives: taint remains tied to the exact units of illicit origin. This work bridges the gap between regulatory needs and technical feasibility by delivering a provably correct, efficient mechanism for tracking tainted cryptocurrency funds at a granular level
Optimizing Cloud Virtual Machine Migration: Minimizing Downtime and Migration Time Using Machine Learning
No full textCloud computing has revolutionized the way services are delivered to users, offering unparalleled flexibility and scalability. However, cloud services can become temporarily unavailable due to maintenance, resource allocation, load balancing, cyberattacks, power management, fault tolerance, and various other factors. To ensure a seamless experience for clients during these periods, live virtual machine migration (LVM) emerges as an indispensable choice. LVM involves relocating virtual machines (VMs) from source to destination with minimal disruption to client activities. While the pre-copy method is the most commonly used live migration technique due to its reliability, it faces challenges such as extended downtime and migration time caused by a large number of dirty pages generated in each iteration. To optimize these migration metrics, numerous solutions have been developed. Nevertheless, a recurring issue in most techniques is the use of static stopping conditions. If the dirty rate exceeds the network throughput, the number of dirty pages retransmitted in each iteration increases, and the hypervisor cannot complete the dirty page transfer in a specific iteration. Moreover, there is no universal stopping condition suitable for all VMs. As a result, the source VM needs to be suspended for a longer time to complete the migration, increasing downtime. Extended migration downtime causes service interruptions and affects the performance of running applications. Therefore, optimizing the memory dirty rate to minimize downtime and total migration time is the primary challenge in the pre-copy approach. To address these challenges, we conducted a thorough analysis of the critical factors influencing live migration performance. Subsequently, we devised an algorithm to identify these critical features and leveraged them to build a machine-learning model capable of intelligently predicting the optimal time to transition into the stop and copy phase, reducing reliance on static stopping conditions. Our proposed machine learning method was rigorously evaluated through experiments conducted on a dedicated testbed using KVM/QUEM technology, involving different VM sizes and memory-intensive workloads. A comparative analysis against proposed pre-copy methods and existing techniques reveals a remarkable improvement, with an average 61.24% reduction in downtime for different RAM configurations in high-write-intensive workloads, along with an average reduction in total migration time of approximately 85.81%. Furthermore, we examined the security concerns surrounding live migration, particularly in domains handling critical applications such as banking, healthcare, etc. Many organizations
in these sectors are hesitant to employ live migration due to security risks. To address this, we introduced a selective encryption approach for protecting sensitive information during migration. Our experimental results highlight that the selective encryption method enhances our proposed machine learning model to reduce downtime and total migration time while preserving the privacy of sensitive data
Contributions to configurable business process management : a knowledge-based , fragmentation, and entropy measurement approach
No full textNous proposons une approche de gestion des processus métier configurables (cBP) offrant une vue consolidée des variantes de ces processus. Une variante est le résultat de l'ajustement d'un cBP en réponse à des besoins fonctionnels et / ou structurels. Cette approche utilise une base de connaissances pour suivre les spécificités de chaque variante représentée sous la forme d'arborescence de structure de processus configurable (cPST). Une implémentation de l'approche consistant à générer un cPST à partir d'un cBP est proposée. Nous appliquons cette approche pour calculer la complexité des processus métier configurables et prouver son avantage pour améliorer la qualité du cBPWe propose an approach for managing configurable Business Processes (cBP) that offer a consolidated view over these processes’ variants. A variant is the result of adjusting a BP in response to functional and/or structural needs. The approach uses a knowledge base to track the specificities of each variant that is represented as configurable Process Structure Tree (cPST). An implementation of the approach that consists of generating a cPST from a cBP is proposed. We apply this approach to compute configurable business process complexity and to prove its advantage for improving cBP qualit
Knowledge Injection from a Domain Sentiment Ontology in an Attention Neural Network for Aspect-Based Sentiment Classification
No full textOver the last years, enormous amounts of opinions have become available on the Web, which causes the interest in the task of Aspect-Based Sentiment Classification (ABSC) to rise. Hybrid models for ABSC, which combine a knowledge base with a machine learning algorithm, have gained popularity because of their superior performance. LCR-Rot-hop-ont++ is such a hybrid model, which injects knowledge from a domain sentiment ontology into the well-performing attention neural network LCR-Rot-hop++. In this work, we extend the LCR-Rot-hop-ont++ model in two ways. First, we inject additional knowledge from a domain sentiment ontology into the neural network. Second, we apply a novel weighting mechanism to the injected tokens to control the influence of additional knowledge. Using the SemEval 2015 and SemEval 2016 datasets for evaluation, we find that knowledge injection improves accuracy for datasets with a limited number of observations. Furthermore, we find that the proposed weighting mechanism leads to improved predictive performance of the neural network model.</p
Talking buildings: Interactive human-building smart-bot for smart buildings
Get PDFConventional conversational agents have a limited ability to respond to different user intents when interacting with smart buildings. The uniqueness of each building, coupled with the heterogeneity of built environments, makes it challenging to adapt communication methods universally. A possible solution is to develop a conversational agent capable of understanding physical, logical, and virtual assets in the built environment, with the aim of establishing a standardised method of human-building communication. Current smart building ontologies and metadata description schemas aim to give smart buildings a common language for the rapidly growing number of devices in smart buildings. This research paper focuses on developing a comprehensive smart building framework that integrates chatbot-driven natural language interactions into smart buildings using the SPARQL query language to query the smart building Knowledge Base (KB) and interact with buildings using a chatbot. An environmental sensor network testbed was set up using an ontology representing a smart building to evaluate the answer to the question. We have used transformer-based machine learning (ML) models to translate the natural language (NL) queries into SPARQL queries and summarise combined SPARQL and natural language queries, which produced promising performance. By integrating chatbots into smart building systems on the edge, users can interact in natural language, provide real-time information, and detect potential threats without the need for specialised knowledge. Our future work will be to extend this model to support heterogeneous building types represented by smart building ontologies. The source code and data sets are publicly available. https://github.com/suhasdevmane/abacws-chatbo
Explaining a Deep Learning Model for Aspect-Based Sentiment Analysis Using SHAP
No full textState-of-the-art machine learning models have continuously improved over recent years, leading to increasingly high-performing models. Simultaneously, it has become increasingly difficult to infer insights in model behavior, as models continuously increase in complexity. This paper aims to increase the explainability of the HAABSA++ model, a state-of-the-art machine learning algorithm that combines a domain sentiment ontology with a deep learning model using contextualized word embeddings, designed for aspect-based sentiment analysis. The model is trained and evaluated on the SemEval 2016 restaurant reviews dataset. For model explainability we propose two SHAP approaches. The first approach, SHAP model 1, applies SHAP after BERT word embeddings are generated, measuring the contribution of each embedded word to the sentiment prediction. The latter approach, SHAP model 2, applies SHAP before BERT word embeddings are generated. SHAP model 2 appears to be better at finding words that individually contribute most towards a sentiment prediction since it generates new word embeddings for each subset of the original sentence.</p
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
- …
