MOBMU 2025 Program
Excel to HTML
WEDNESDAY 5 FEBRUARY 2025
Emerging Technologies
Session Chair: David Akopian, University of Texas, San Antonio
09:30 - 10:30
Harbour A
09:30MOBMU-304
Financial risk detection: A zero-shot approach to identifying insider trading in earnings reports, Klaus Schwarz, University of Granada (Spain); Kendrick Bollens, SRH University of Applied Sciences (Germany) [view abstract]
This paper presents a zero-shot classification approach to detect potential insider trading risks within financial disclosures and earnings reports. Insider trading can compromise market fairness, yet traditional detection methods often rely on retrospective analysis and manually labeled data. To address this, our approach leverages advancements in natural language processing (NLP), utilizing a pre-trained language model for zero-shot classification. By analyzing financial documents for patterns like unusual transactions or significant corporate changes, the approach efficiently identifies potential risks without the need for domain-specific training data. We validate its effectiveness by comparing results against known insider trading cases, showing its potential as a tool for early risk detection. This scalable, low-resource method offers valuable applications for financial compliance and market surveillance, with the flexibility to adapt to other domains.
09:50MOBMU-305
Scraping, clustering, ranking, topic modeling - A case study of technology news aggregation, Aleksandr Rykov, SRH University of Applied Sciences (Germany); Francisco Saavedra, SRH Berlin University of Applied Sciences; Klaus Schwarz, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, technische Hochschule Brandenburg (Germany) [view abstract]
Today's world is characterized by the high volume of content generated, including news media. Finding reliable, relative, and well-written articles to stay informed can be time-consuming. Here is where the news aggregation comes in place. It simplifies the process and allows for selecting the most relevant news according to specific preferences while keeping all the relevant information in one place. Such a system consists of various components: web scraping from news sources, structuring obtained data, ranking and scoring texts, extracting semantics with text embeddings, clustering, and topic modeling.Clustering and ranking news articles is not a new problem, and various solutions already exist. This study explores an approach that combines traditional methods with AI techniques. Initially, articles were downloaded and then processed using a combination of tests and metrics. BART, GTE, and LLM models were used to analyze and calculate embeddings. The dimensionality of embeddings was then reduced with UMAP dimensionality reduction, followed by clustering with HDBSCAN and labeling using LLaMA.The resulting dataset consists of 431 clusters and 6,675 scored and ranked articles, with approximately 13.36% identified as outliers not belonging to any cluster.
10:10MOBMU-306
Balancing expert assessments with AI acceleration to foster foresight at Siemens professional education, Stephan Szuppa, SRH Berlin University of Applied Sciences (Germany); Rene Rohrbeck, EDHEC Business School (France) [view abstract]
The digital transition in industry is based on the technological competencies of involved experts. Artificial Intelligence (AI) is anticipated to accelerate foresight, particularly in complex environments. So, integrating data-driven insights with expert assessments is crucial for enhancing decision-making. This study examines the current state and challenges of AI-assisted foresight through a particular case: Siemens Professional Education (SPE), a Human Resources organization of Siemens AG responsible for planning and facilitating vocational training, dual higher-education programs, and reskilling and upskilling initiatives for over 320,000 employees globally. The foresight approach for SPE must account for present and future technology and business trends across three verticals, workforce reskilling demands and evolving job market needs. We discuss SPE's proven foresight methodology, critically review the AI impact and its benefits, and contextualize it within the emerging literature on foresight automation and AI integration. Finally, the outcome of SPE Innovation Management must be very concrete and support Siemens Businesses in 3 verticals: Digital Industry, Infrastructure and Mobility. Currently, education content for the following high prior Trend cluster is under development and the required competency maps can be shared exemplary: Responsible Business Usage of AI, Decentralized Energy and Storage (continued) Circularity and Shortage of Resources Resilience of Systems and Critical Infrastructure Entrepreneurial Mindset for Tomorrow's Workforce The study concludes by proposing a comprehensive framework and identifying critical success factors for balancing expert assessment with the application of AI support in strategic foresight.
10:30 – 11:00 Coffee Break
Imaging & Multimedia & Mobile Apps
Session Chair: David Akopian, University of Texas, San Antonio
11:00 - 12:20
Harbour A
11:00MOBMU-307
Interactive posters motivated by VFX and 360-degree video technology: Enhancing MICE experiences in a German leisure region, Dirk Hagen, Hochschule Hannover - University of Applied Sciences and Arts (Germany); Eberhard Hasche, TH Brandenburg (Germany); Reiner Creutzburg, SRH Berlin University of Apploied Sciences (Germany) [view abstract]
The rapid advances in imaging technology, particularly in visual effects (VFX) and 360-degree video, can significantly transform the MICE (Meetings, Incentives, Conventions, and Exhibitions) industry. These technological innovations are of great importance for creating more immersive environments that can significantly enhance the engagement and satisfaction of event participants. This article examines the implementation of these technologies through interactive posters as part of experience trails in L�neburg Heath, a well-known leisure region in northern Germany.The deployment of 360-degree panoramic videos and VFX for interactive posters in specially equipped conference rooms, such as the "Dortmund" conference room at the Park Soltau Hotel (Germany), exemplifies the capacity of these technologies to transform traditional business meeting environments. The paper will present the technological approach of such imaging technologies, which are necessary to create such interactive posters successfully and discuss possible limitations and further potential in this case study.The unique selling point of VFX and 360-degree video technologies is their ability to combine business and leisure elements. This feature has the potential to significantly enhance the return on investment for stakeholders and organizers. However, this competitive advantage is not without its challenges, including high setup costs and the need for specialized content production.A key value proposition is the potential of VFX and 360-degree video technologies to enrich the experience of event participants. By facilitating the delivery of more immersive content, these technologies enhance participant engagement and empower venues such as conference hotels to distinguish themselves in a highly competitive environment, benefiting various stakeholders.
11:20MOBMU-308
iForester, an iOS app for automated tree inventory., Zhiheng Yin, Purdue University (US); Wang Xiang; Songlin Fei; Song Zhang, [view abstract]
Tree trunk diameter measurement, particularly Diameter at Breast Height (DBH), is essential in tree inventory and forest management. Traditionally, DBH has been measured using diameter tapes, a labor-intensive method prone to human error and challenging in inaccessible areas. As digital forestry advances, mobile technology offers promising solutions for DBH data collection. However, existing methods such as AR-based systems face challenges with setup complexity and inconsistent accuracy under varying environmental conditions. Laser-scanning approaches, while accurate, require large storage space, consume excessive battery power, and are susceptible to environmental interference. A clean trunk segmentation is critical for accurate DBH calculation. Recognizing these limitations, we developed an efficient and fast tree trunk segmentation method using 3D Lidar data captured via iPhone devices and released the iForester iOS app to facilitate field testing. Our approach minimizes storage needs, improves battery efficiency, and ensures robust performance across different conditions. By producing clean tree trunk segmentation, it enables highly accurate DBH measurements. This method bridges the gap between advanced algorithms and practical, real-world applications, supporting the future of digital and precision forestry.
11:40MOBMU-309
The potential and limitations of computer vision in plant stress detection, Lukasz Rojek, SRH Berlin University of Applied Sciences (Germany) [view abstract]
Computer vision technologies are increasingly applied across diverse fields, from object detection to complex analytical systems. In agriculture, methods such as thermal and short-wave imagery enable monitoring plant conditions through indices such as the Normalized Difference Water Index (NDWI) and Crop Water Stress Index (CWSI). This paper explores the potential and limitations of applying computer vision for plant stress detection, providing insights to advance understanding in this growing field.
12:00MOBMU-310
Detecting voice cloning and text to speech audio in real time on mobile devices, Waldemar Berchtold, Fraunhofer SIT (Germany); Julian Heeger, Fraunhofer SIT (Germany); Simon Bugert, Fraunhofer SIT (Germany); Martin Steinebach, Fraunhofer SIT (Germany) [view abstract]
Recent incidents involving identity attacks through audio manipulation have become increasingly prevalent. For instance, the New York Post reported on an incident where a 15-year-old girl's voice was cloned in an extortion attempt, with the perpetrators demanding a million-dollar ransom from her mother. A study conducted by McAfee highlights the growing threat, revealing that 25% of respondents or their acquaintances had experienced AI-driven audio fraud. Of the people who reported losing money, 36% said they lost between $500 and $3, 000, while 7% got taken for sums anywhere between $5, 000 and $15, 000. In this paper, we present a method that analyzes an audio stream in real time and provides an indication of whether it is an identity attack or not.
Mobile Devices and Multimedia: Enabling Technologies, Algorithms, and Applications Posters (with lunch)
12:20 - 14:00
The Grove
12:20MOBMU-311
A blockchain-based digital voting solution for securing democracy, Cristian Cristian , SRH Berlin University of Applied Sciences (Germany); Navaneeth Shivananjappa, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin (Germany) [view abstract]
The main objective of this paper is to understand what and how the process of deploying a digital voting system in a democratic country would look like and what challenges a potential architect of such a system might confront. More than that, the study will focus on using blockchain to solve specific barriers that might appear in the quest for a digitized way of voting. By exploring the confluence of technology and democracy, this research underscores the importance of technological innovation in democratic participation and governance. This paper will cover the fundamentals of blockchain, its integration within electronic voting systems, examination of its technical attributes, societal impacts, regulatory compliance, and a proposed blockchain-based e-voting system. Certain features of blockchain present a promising solution to longstanding challenges in electronic voting systems, such as ensuring security, integrity, compliance, and public trust. The initial sections of the paper provide an in-depth analysis of blockchain's theoretical basis, emphasizing how its unique attributes can safeguard the electoral process against fraud and manipulation. Even more so, security techniques like cryptography and consensus mechanisms are also studied in this section. Following this inquiry, the study investigates the societal implications of deploying an e-voting system in a democracy, drawing insights from the experiences of various European countries. This exploration reveals the critical role of societal acceptance, efficient technical solutions, and legal frameworks in successfully implementing new voting methods. Next, the design phase articulates a viable system architecture that balances several components: efficiency, control, security, and transparency. The suggested architecture considers critical factors such as voter anonymity, vote integrity, and the unique ability to modify votes within the voting period while aligning with democratic principles and regulatory standards. In summary, this paper presents a complete analysis of digital voting, including technical and societal considerations, steps to deploy such a system in a democratic country, and, more importantly, blockchain technology's potential to revolutionize it further. Integrating theoretical knowledge, societal insights, and practical design considerations offers a blueprint or framework for future advancements in creating digital democracies.
12:20MOBMU-312
Cybersecurity awareness among young adults: An analytical study, Mahipal Mahipal, SRH Berlin University of Applied Sciences (Germany); Navaneeth Shivananjappa, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin University of Applied Sciences (Germany) [view abstract]
This research investigates the cybersecurity awareness of young adults aged 18 to 27 through a structured survey. It evaluates key areas such as phishing recognition, two-factor authentication usage, data backup practices, and knowledge of encryption. The findings indicate that while many participants actively use two-factor authentication, there are notable gaps in understanding encryption and managing privacy settings. Differences in awareness were observed between age groups, with younger individuals displaying lower confidence in technical skills. The results highlight the importance of targeted educational programs to address these knowledge gaps and enhance online safety practices. These initiatives are essential to help young adults strengthen their cybersecurity defenses and protect their personal information in an increasingly digital world. The study underscores the continuous need for education to stay ahead of evolving cyber threats.
12:20MOBMU-313
Overview and state-of-the-art in cybersecurity threat live maps, Remilekun Adeopatoye, SRH Berlin University of Applied Sciences (Germany); Merabatte Eddine, SRH Berlin University of Applied Sciences (Germany); Asem Mousa, SRH Berlin University of Applied Sciences (Germany); Hamroz Gavharov, SRH berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin University of Applied Sciences (Germany) [view abstract]
The necessity for real-time situational awareness in cybersecurity is becoming critical in a rapidly connected world. Live Threat Maps was introduced as a dynamic visualization tool for real-time monitoring and analysis of security threats. It uses data from network traffic, intrusion detection systems, malware reports, and threat intelligence feeds to provide a comprehensive overview of the evolving threat landscape. The paper evaluates and discusses the design principles, architectural considerations, and integration of machine learning algorithms and anomaly detection techniques. It also discusses the practical applications of Live Threat Maps for security operations centers, incident response teams, and decision-makers. The paper addresses data accuracy, scalability, and privacy concerns.
12:20MOBMU-314
Automated monitoring stolen cultural artifacts on online marketplaces, Huajian Liu, Fraunhofer SIT (Germany); York Yannikos, Fraunhofer SIT (Germany); Julian Heeger, Fraunhofer SIT (Germany); Simon Bugert, Fraunhofer SIT (Germany); Waldemar Berchtold, Fraunhofer SIT (Germany); Martin Steinebach, Fraunhofer SIT (Germany) [view abstract]
Tracking and identifying stolen cultural artifacts on online marketplaces is a daunting task that has to be accomplished through manual search. In this paper, an automated monitoring tool is developed to track and identify stolen cultural goods on targeted online sales platforms. In case of theft, the original owner can upload descriptive keywords and photos of the stolen objects to start monitoring tasks to track and identify the stolen objects on targeted online marketplaces and get alerted when identical or highly similar objects appear on the monitored sales platforms. The technical challenges posed by automated monitoring are addressed by proposed advanced crawling and image feature extraction and matching solutions. With the support of proposed novel techniques, the developed monitoring tool can efficiently and effectively monitor stolen artifacts on online marketplaces, significantly reducing the manual inspection effort.
12:20MOBMU-315
A review of cybersecurity and challenges associated with microgrid (MG) systems, Saiful islam, SRH Berlin University of Applied Sciences (Germany); klaus schwarz, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin University of Applied Sciences (Germany); Michael Hartmann, SRH Berlin University of Applied Sciences (Germany) [view abstract]
Microgrid systems face several challenges, including critical issues such as load scheduling and the generation of large amounts of data that need to be monitored to stabilize the system during operation. It is crucial to secure the real-time collection of data and protect it throughout the process. In this context, infrastructure such as cloud computing and data security plays an important role in overcoming the situation. This paper will provide a critical analysis of current solutions and future perspectives on data sharing in microgrids and their challenges.
12:20MOBMU-316
An energy management system to address problems and challenges associated with microgrids, Saiful islam, SRH Berlin University of Applied Sciences (Germany); klaus schwarz, SRH Berlin University of Applied Sciences (Germany); Michael Hartmann, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin University of Applied Sciences (Germany) [view abstract]
Microgrids are associated with various problems and challenges, such as energy shortages, overcharging, excessive grid purchases, and more. Additionally, technical issues such as frequency distortion and losses in the equipment for connected DERs can be addressed by assigning an Energy Management System (EMS) as a control strategy.
12:20MOBMU-317
San Antonio Research Partnership Portal: Smart artificial intelligent community applications, Mohammad Rokim, Department of Urban and Regional Planning, Khulna University of Engineering & Technology (Bangladesh); Mohammad NadimCenter for Cybersecurity Innovation, Texas A&M University - Central Texas (US); David Akopian; Department of Electrical & Computer Engineering, The University of Texas at San Antonio (US); Adolfo Matamoros, School of Civil & Environmental Engineering, The University of Texas at San Antonio (US) [view abstract]
Urban governance is vital for efficiently managing cities, promoting sustainable development, and improving quality of life for residents. In the realm of urban governance, the San Antonio Research Partnership Portal stands as a groundbreaking initiative, fostering collaboration between diverse city entities and leveraging innovative smart applications. In this paper, we will focus on its ability to facilitate strategic alignment among city departments, public feedback integration, and streamlined collaboration with academic institutions. Through technical insights and real-world case studies, this paper underscores the portal's role in enhancing municipal responsiveness, improving decision-making processes, and exemplifying the potential of smart applications utilizing artificial intelligence for fostering effective city management and community engagement.
12:20MOBMU-318
A characterization of chatbot development platforms for deep-logic operations, Ensieh Modiridovom, University of Texas at San Antonio (US); David Akopian [view abstract]
12:20MOBMU-319
Integration of task-oriented chatbots with generative AI, Hasmik Yengibaryan, Yerevan State University (Armenia); David Akopian, University of Texas at San Antonio (US) [view abstract]
Nowadays, chatbots have become a pivotal part of our daily life enhancing human-machine interaction across various domains. They are tools that are reshaping business-customer interactions. From customer service to healthcare, education to finance, chatbots are increasingly important in various industries. Chatbots, or conversational agents, are computer programs designed to simulate human conversation with users, usually through text or voice interactions.Platforms such as Amazon Lex, IBM Watson, ManyChat, Chatfuel, and Google Dialogflow have played a significant role in advancing this technology, mostly focusing on expert-protocol-driven chatbots to configure the human-machine conversation. However, advancements in Generative AI resulted in more flexible, contextually aware, and human-like chatbots.Despite the impressive capabilities of expert-protocol-driven platforms and generative AI systems, a notable gap exists between these approaches. While expert-driven platforms offer robust solutions for specific tasks through structured interactions, they often lack the broader conversational flexibility and contextual richness found in generative AI. This paper explores the concept of integrating task-oriented chatbots with generative AI.It demonstrates this concept by examining the integration of Dialogflow and Gemini to bridge the gap between task-oriented and generative chatbots. A case study integration MachineQuizzing chatbot is developed for an enhanced learning experience.
12:20MOBMU-320
Enhancing brain tumor detection: Leveraging convolutional neural network (CNN) models for improved diagnostic accuracy, Mahnoor Jamil, Student (Macedonia); Reiner Creutzburg, Professor (Germany) [view abstract]
Brain tumor detection is a critical component of medical diagnostics, aiming to provide accurate, timely identification of tumor presence. This study utilizes a Convolutional Neural Network (CNN) approach with the VGG-16 model architecture to classify brain MRI images as either showing the presence of a tumor or not. Leveraging transfer learning, VGG-16 was fine-tuned for binary classification on a dataset of brain MRI images. This approach achieved validation and test accuracies of approximately 88% and 80%, respectively. Our methodology combines image preprocessing techniques with data augmentation to enhance model robustness on limited datasets. The results demonstrate the potential of CNN-based deep learning models in automated medical imaging and suggest future improvements through dataset expansion and model fine-tuning
12:20MOBMU-321
Enhancing health promotion communication through domain-tailoring techniques in ChatGPT, Ebrahim Mellatdoust Pordel, University of Texas at San Antonio (US); Farin Ahmed, The University of Texas at San Antonio (US); David Akopian, The University of Texas at San Antonio (US) [view abstract]
With the exponential growth of large language models (LLMs), enhancing model adaptability for diverse real-world applications has become crucial. This study critically examines domain-specific fine-tuning of ChatGPT, highlighting the delicate balance between generalizability and specificity in health promotion communication. Employing two distinct fine-tuning strategies�single-prompt interactions and multi-turn conversation models�the research advances current methodologies for tailoring LLMs to specialized domains. By incorporating approaches such as data augmentation, transfer learning, and adaptive fine-tuning, the study systematically evaluates ChatGPT's adaptability in handling health-specific dialogues, comparing model performance across varied interaction types. Case studies and targeted customization strategies underscore the practical utility and significant impact of these adaptations in applied health communication contexts, demonstrating the enhanced contextual understanding in multi-turn interactions. Results indicate the superior efficacy of the multi-turn approach in managing nuanced, contextually rich dialogues, underscoring the capacity of the model for sustained engagement in health-related discourse. These findings have significant implications for advancing AI in health communication, suggesting a developmental trajectory that integrates technological sophistication with a focus on empathetic user engagement.
12:20MOBMU-322
Contrast-based visual inspection for automated deburring in manufacturing: A cobot approach, Yash Sharma, SRH Berlin University of Applied Sciences (Germany); Lukasz Rojek, SRH Berlin University of Applied Sciences (Germany) [view abstract]
This paper uncovers a new concept in automating deburring processes in manufacturing industries through a cobot-based contrast perception inspection method to improve the quality and reliability of manufacturing plants. Therefore, the existing conventional deburring methods involve the inspection process, in which variability arises due to many factors, such as the fatigue of the inspector. In this paper, an analysis of using a cobot with a Baumer Veri Sense camera for visual inspection, providing a high contrast, stable image view of roller parts for burr detection, is discussed. Turner machine is incorporated with Finder relay along with Universal Robot for deburring part where cobot captures images of rollers both front and rear side to recognize burrs for further deburring followed by polishing with spiritus-based machine. The described method of roller deburring automation also eliminates the need for workers' involvement, improves quality and enhances time and cost efficiency.To sum up, the given studies show rather promising results compared to the traditional manual deburring. However, the research has definite drawbacks: it uses only one type of roller and is based on limited experimentations involving only one hundred rollers. Therefore, further development of the given approach to deburring different materials and components is required. The present work gives some knowledge about the functionality of cobots and suggests how contrast-based visual inspection can be used in highly reliable industrial applications.
12:20MOBMU-323
Overcoming platform limitations: A standalone task-oriented dialog system support, Ahmer Patel, The University of Texas at San Antonio (US); David Akopian, The University of Texas at San Antonio (US); Sushil Prasad, The University of Texas at San Antonio (US) [view abstract]
This paper presents the design and development of a standalone, cross-platform chatbot system that addresses the limitations imposed by mainstream messaging platforms. While platforms like Facebook Messenger and WhatsApp facilitate chatbot deployment, they impose restrictions such as message frequency, timing, and user access. These limitations disrupt continuous, deep-logic interactions with users, hindering the effectiveness of extended engagement in customer service and support. This project proposes a solution using Google Dialogflow for natural language processing (NLP), a Cross-Platform client application for seamless cross-platform support, and Firebase for backend data storage and real-time messaging. The system supports message continuity, multimedia features, quick-reply buttons, and scalability for up to 10,000 users. Through a flexible, independent design, this prototype demonstrates a scalable, unrestricted chatbot solution for improved user satisfaction and engagement across devices. We aimed to make the application user-friendly, scalable, and safe.
15:00 – 15:30 Coffee Break
Machine Learning and AI
Session Chair: David Akopian, University of Texas, San Antonio
15:30 - 16:10
Harbour A
15:30MOBMU-324
AI-powered perception systems for intelligent autonomous vehicles (IAV), Nithin Jayagovindan, SRH University of Applied Sciences (Germany); Alexander Iliev, SRH University of Applied Sciences (Germany); Reiner Creutzburg, TH Brandenburg (Germany) [view abstract]
Artificial Intelligence (AI) is pivotal in advancing autonomous vehicles, transforming them into safer, more intelligent, and reliable transportation systems. This paper explores state-of-the-art AI techniques and algorithms for implementing Intelligent Autonomous Vehicles (IAVs), focusing on enhanced perception, decision-making, and control. Key contributions include the integration of deep learning frameworks like YOLOv8 for real-time traffic detection and ENet for semantic segmentation, enabling accurate and robust traffic lane detection, pedestrian recognition, and traffic sign classification across diverse and dynamic environments. While these advancements address critical challenges such as varying weather conditions and dynamic traffic scenarios, the increasing reliance on AI highlights cybersecurity concerns. Autonomous vehicles are inherently vulnerable to threats like data tampering, adversarial attacks on AI models, and unauthorized access to critical systems. To mitigate these risks, this paper emphasizes the importance of integrating cybersecurity measures into AI-driven frameworks. Secure data pipelines, robust encryption mechanisms, and adversarial resilience techniques are proposed to safeguard vehicle and passenger safety. The proposed framework showcases technological innovations and addresses the intersection of AI and cybersecurity, ensuring the reliable, secure, and efficient operation of IAVs. By optimizing vehicle dynamics through secure AI-based control systems, this research advances the vision of intelligent, sustainable, and accessible mobility solutions. These contributions mark a significant step toward realizing intelligent transportation systems while mitigating cybersecurity risks, paving the way for a future where autonomous vehicles become integral to both sustainable mobility and secure digital infrastructure
15:50MOBMU-325
CommunityInsight AI: A community-centric urban governance application driven by AI, Mohammad Rokim, Department of Urban and Regional Planning, Khulna University of Engineering & Technology (Bangladesh); Mohammad Nadim, Center for Cybersecurity Innovation, Texas A&M University - Central Texas (US); David Akopian, Department of Electrical & Computer Engineering, The University of Texas at San Antonio (US); Adolfo Matamoros, School of Civil & Environmental Engineering, The University of Texas at San Antonio (US) [view abstract]
In the era of data-driven decision making, cities and communities are increasingly seeking ways to effectively gather insights from public feedback and comments to shape their research and development initiatives. Town hall community meetings serve as a valuable platform for citizens to express their opinions, concerns, and ideas about various aspects of city life. In this study, we aim to explore the effectiveness of different keyword extraction tools and similarity matching algorithms in matching town hall community comments with city strategic plans and current research opportunities. We employ KPMiner, TopicRank, MultipartiteRank, and KeyBERT for keyword extraction, and evaluate the performance of cosine similarity, word embedding similarity, and BERT-based similarity for matching the extracted keywords. By combining these techniques, we aim to bridge the gap between community feedback and research initiatives, enabling data-driven decision-making in urban development. Our findings will provide valuable insights for more inclusive and informed strategies, ensuring that citizen opinions and concerns are effectively incorporated into city planning and development efforts.
Topics in Cybersecurity
Session Chair: David Akopian, University of Texas, San Antonio
16:10 - 17:10
Harbour A
16:10MOBMU-326
AI-based vulnerability scanners: A cross-sectional survey analysis, Sam Chemparathy, SRH Berlin University of Applied Sciences (Germany); Navaneeth Shivananjappa, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, SRH Berlin University of Applied Sciences (Germany) [view abstract]
One major innovation in improving organizations' security measures is the adoption of AI-based vulnerability scanners within the cybersecurity space. The paper analyzes cross-sectional survey research identifying factors that influence the acceptance and use of such advanced tools among cybersecurity professionals. The primary method of gathering data was a structured survey questionnaire that used Likert-scale questions to quantify the participants' opinions objectively. It contained 20 questions based on established models, including TAM, UTAUT, and IDT. In this research, the total number of people who responded to the survey was 49, comprising cybersecurity professionals working in various industry domains. This instrument has measured perceived usefulness, ease of use, performance expectancy, effort expectancy, social influence, facilitating conditions, and the stages of adoption, including awareness, interest, evaluation, trial, and adoption. Our results provide insight into factors that drive or hinder the adoption of AI-based vulnerability scanners, focusing on the significant role of perceived benefits and organizational support. The present paper offers valuable implications for practitioners and researchers who aim to foster AI-driven security solutions within organizational contexts.
16:30MOBMU-327
OSINT investigation of social media for classifying reactions on cannabis legalization, Berfin Atabay, SRH Berlin University of Applied Sciences (Germany); Resul G�m�s, SRH Berlin (Germany); Klaus Schwarz, SRH Berlin University of Applied Sciences (Germany); Reiner Creutzburg, TH Brandenburg (Germany) [view abstract]
Cannabis legalization is an important socio-political issueworldwide, creating a variety of public opinions and sentiments.This study uses Open Source Intelligence (OSINT) techniquesand methods to analyze social media opinion around the recentlegalization of Cannabis in Germany on April 1, 2024, specificallyon the Reddit platform. By collecting and analyzing commentsfrom Reddit, we aim to group and understand public sentimenton this topic. Our methodology includes data collection,cleaning, clustering, and sentiment analysis using natural languageprocessing techniques. We use a sentence transformermodel for text embedding and the K-means clustering algorithm to find different angles around legalization. Sentiment analysiswas implemented using a pre-trained model developed byHugging Face based on BERT. The results reveal seven distinctclusters representing different themes and sentiments aboutcannabis legalization. Overall, the sentiment and public opiniontend to lean positive, indicating general public support for Cannabis legalization despite some expectations. In addition todemonstrating the value of OSINT in social media research, thisstudy offers insight into public sentiment surrounding Cannabislegalization to researchers and policymakers.
16:50MOBMU-328
PrivacyBuddy: An Android privacy dashboard for detecting excessive data collection with a focus on location data, Toon Dehaene, KU Leuven DistriNet (Belgium); Maxime Bellis, KULeuven (Belgium); Tristan Pelgrims, KU Leuven (Belgium); Vincent Naessens, KU Leuven DistriNet (Belgium); Bert Lagaisse, KU Leuven DistriNet (Belgium) [view abstract]
This paper presents PrivacyBuddy, an innovative Android privacy dashboard designed to detect excessive data collection, focusing on location data. It addresses the growing concern over apps that collect more personal information than necessary, often leading to privacy violations. Current Android privacy dashboards are inadequate, lacking effective visualizations and insights into app behavior. PrivacyBuddy enhances user awareness by providing intuitive visualizations of data collection frequency, volume, and type, along with a Privacy Score for each app. The dashboard employs user-centered design principles to cater to diverse user needs, from casual users to privacy-conscious individuals. Key features include detailed tracking of location requests, distinguishing between foreground and background access, and offering actionable insights for users to manage their privacy settings effectively. A secure implementation extends the Android OS to monitor data requests made by other apps. A user study validates the design, demonstrating improved usability and user engagement compared to existing solutions. This work contributes to the field of visual data analytics by merging design principles with practical applications, ultimately empowering users to take control of their data privacy in an increasingly data-driven world.