In today’s data-driven business landscape, the need for advanced analytics solutions is paramount for informed decision-making and sustainable growth. This paper introduces DataInsight Big Data Analytics Services, a comprehensive system tailored to meet the analytical demands of businesses, particularly in handling big data. DataInsight integrates front-end and back-end technologies, machine learning algorithms, and various data analytics techniques to provide business owners with actionable insights into the future of their enterprises. Our approach focuses on developing a robust system for business data analytics by carefully selecting algorithms that are capable of handling large-scale data and are highly beneficial for business applications. A significant contribution to this research is the improvement of Winnow algorithm and adding feature selection methods for more accurate results. Another contribution is the generation of experimental data to evaluate the performance of the core algorithms used in this system, demonstrating the effectiveness and versatility of the DataInsight system. Three big data analytics methods for Clustering, Association Rule Mining, and Classification, were examined through multiple tests to ensure their capabilities and their effectiveness.

Federated Learning (FL) is a machine learning framework that allows multiple clients to contribute their data to a single machine learning model without sacrificing their privacy. Although FL addresses some security issues, it is still susceptible to model poisoning attacks where malicious clients aim to corrupt the main learning model by sending poisoned updates. Byzantinerobust methods are defenses that aim to prevent corruption of the main learning model by tolerating a certain number of malicious clients. However, they can only resist a small number of malicious clients. This proposed method uses Gap statistics to determine the optimal number of clusters to cluster clients. This will improve the detection accuracy of malicious clients while preventing the mislcassification of honest clients in a Federated Learning setting. Our experiments so far show us an improvement over the base method.

In this paper, we present Random Network Byzantine Fault Tolerance (RNBFT), a novel, partially synchronous, Byzantine Fault Tolerance (BFT) consensus protocol aimed towards large-scale consortium blockchain networks.The essence of RNBFT lies in the random network communication paradigm which reduces the communication overhead of the system and is further enhanced by leveraging aggregation of multi-signatures backed with an optimized gossip paradigm. This approach collectively results in achieving high throughput and efficiency which can be scaled easily with large-size quorums. With a series of experiments and analysis, we affirm that RNBFT is an ideal choice for large-size consortium networks. Thus, RNBFT promises resiliency over both major and minor arbitrary failures in nodes with a fair trade-off between the performance and scalability of the system.

Forming teams of experts who collectively hold a set of required skills and can successfully cooperate is challenging due to the vast pool of feasible candidates with diverse backgrounds, skills, and personalities. Neural models have been proposed to address scalability while maintaining efficacy by learning the distributions of experts and skills from successful teams in the past in order to recommend future teams. However, such models are prone to overfitting when training data suffers from a long-tailed distribution, i.e., few experts have most of the successful collaborations, and the majority has participated sparingly. In this paper, we present an optimization objective that leverages both successful and virtually unsuccessful teams to overcome the long-tailed distribution problem. We propose three negative sampling heuristics that can be seamlessly employed during the training of neural models. We study the synergistic effects of negative samples on the performance of neural models compared to lack thereof on two large-scale benchmark datasets of computer science publications and movies, respectively. Our experiments show that neural models that take unsuccessful teams (negative samples) into account are more efficient and effective in training and inference, respectively.

Online user communities exhibit distinct temporal dynamics in response to popular topics or breaking events. Despite abundant community detection libraries, there is yet to be one that provides access to the possible user communities in future time intervals. To bridge this gap, we contribute SEERa, an open-source end-to-end community prediction framework to identify future user communities in a text streaming social network. SEERa incorporates state-of-the-art temporal graph neural networks to model inter-user topical affinities at each time interval via streams of temporal graphs. This all takes place while users' topics of interest and hence their inter-user topical affinities are changing over time. SEERa predicts yet-to-be-seen user communities on the final positions of users' vectors in the latent space. Notably, our framework serves as a one-stop-shop to future user communities for Social Information Retrieval and Social Recommendation systems.

Statistical analysis of health data is an essential task in healthcare. However, existing healthcare systems are incompatible with this critical need due to privacy restrictions. A recently emerged technology called Blockchain has shown great promise for mitigating this incompatibility. In this work, we aim to improve existing secure statistical analysis protocols by leveraging the blockchain technology. We propose a novel method that enables researchers to perform statistical analysis on health data in a privacy-preserving, secure, and precise manner.