In the age of rapid information exchange, understanding public sentiment on social media platforms like Twitter has become essential for various stakeholders. This paper presents a comprehensive analysis of Twitter Offensive Language Detection problem using several robust, powerful and expressive machine learning models from different model families, including Artificial Neural Networks (ANN), Support Vector Machines (SVM), Decision Trees(DT), and Random Forests (RT). The primary objective is to evaluate the effectiveness of each model in classifying tweets as "offensive" or "not-offensive", which has a priceless precaution value for a great many institutions, corporations, and regulatory bodies with priority against Hate Speech Detection -as it is an aggravated form of Offensive Language Detection-. The studied annotated Twitter dataset is majorly utilized in the literature for a classification task including hate speech detection, which is different from our approach prioritizing the detection of all forms of offensive language regardless of hate speech.Upon tweet preprocessing that involves data cleaning (removal of several not-relevant patterns and stopwords), applying tokenization and stemmization ( Porter stemmization), the feature engineering is held in a way that the effects of unigram and bigram features, sentimental polarity-based feature extraction on overall and class-based accuracies are investigated, concluding with top-1500 unigram feature extraction for prevention of curse of dimensionality.Upon proper and well-administered hyperparameter tuning (incl. Early Stopping Method, Grid Search, Halving Grid Search) with the consideration of computation and storage resources, three (3) models (SVM, ANN, DT) reach 96% overall accuracy ratio (except RF with 95%). Overall, the better-grasped class during the learning process is the "Offensive" class (at best, DT with 97% "Offensive" class accuracy), in accordance with the overall class imbalance. However, worthy to note that, SVM outperforms other models (by 4% or 5%) with 94% "Not-offensive" class accuracy.One of the main contributions of the study is, for Twitter Offensive Language Detection, to demonstrate the power of proper, realistic hyperparameter tuning in the existence of sufficiently expressive model family by the same top-notch overall accuracy result of the three robust model families and 1% less overall accuracy of another robust model family.The class-based performance differences suggest an Ensemble approach combining the different robust model families might gain 1% or 2% overall accuracy based on the early literature.The work is made available on GitHub1for ease of use and access.

In this paper, we address the crucial need for effective and efficient Task Failure Prediction in Service Orchestration and Automation Platforms (SOAP).These complex systems, with inherent dependencies, are prone to task failures; which leads to service disruptions and increased costs. To tackle this challenge, we present a robust Task Failure Prediction Framework designed explicitly for SOAP environments. This framework leverages server heartbeats as a key data source for predicting task failures. It adopts a dual prediction approach, offering users the flexibility to choose between the INIT method (a rapid prediction method based on the initial server state) and the EXECUTION method (a comprehensive approach utilizing complete historical resource metrics).This paper makes several significant contributions, including offering a new perspective on the relationship between system resource features and task failure prediction; providing a systematic end-to-end solution for mitigating task failure risks in the SOAP context, including adaptable data collection, analysis, model training, and prediction for both online and offline scenarios; optimizing task failure prediction for specific SOAP use cases through "Task-wise," "Server-wise," and "Task Type-wise" strategies; and contributing to the understanding of task failure prediction and service orchestration automation by addressing some of the existing challenges and gaps in the literature.We confirm the framework’s validity in predicting task failures based on empirical results from use cases. Specifically, in a case study involving two distinct tasks—UCS-1 (a web service-based task) and UCS-2 (a memory-intensive task)—the EXECUTION method, leveraging Random Forest and XGBoost algorithms, achieved a perfect accuracy 1.0000 for predicting failures in UCS-1, with respective accuracies of 0.9580, and 0.9630 for UCS-2. This underscores the power of using a "Task-wise" complete resource usage history for precise failure prediction.We also highlight the inherent trade-off between accuracy and resource efficiency. While the EXECUTION method excels in accuracy, the INIT method offers substantial efficiency gains while maintaining a commendable level of accuracy. This flexibility empowers SOAP users to tailor the framework to their specific requirements and needs.We make our framework’s current implementation and dataset excerpts available on GitHub to support future research efforts1.

Record linkage, the process of identifying records from different datasets that represent the same real-world entity, faces challenges due to data inconsistencies and privacy concerns. To tackle these challenges, various embedding-based techniques have been developed. These methods embed records into a vector space, allowing secure comparison of the embeddings through methods like secure multi-party computation (SMPC). In this study, we introduce a novel deep learning framework for record linkage that leverages the capabilities of pre-trained Sentence Transformers and Contrastive Learning to generate record-level embeddings. The framework centers around the representation of each record as a semantically rich sentence embedding, with matching records being identified based on the similarity of these embeddings.This paper presents three (3) key contributions to enhance record linkage processes while maintaining privacy. First, we demonstrate that raw pre-trained sentence embeddings effectively preserve semantic similarity (with no additional training phase or fine-tuning), with the model STBRM outperforming unsupervised competitors across various datasets. Second, we introduce a Self-Supervised Contrastive Learning method, STBCT, which enables organizations to enhance their privacy by avoiding the sharing of sensitive data for global model fine-tuning and incorporates a local fine-tuning technique to tailor embeddings for the record linkage task. Lastly, to improve computational efficiency and privacy in cryptographic contexts, we propose a Bit-Vectorization Technique, STBBV, which converts learned sentence embeddings into a compact binary format, optimizing their usability in SMPC.We conducted extensive experimental evaluations on four benchmark datasets. The results demonstrate that our proposed framework achieves performance comparable to or better than state-of-the-art methods. Notably, our top-performing model, STBCT, outperforms its main counterparts on average by up to 20% on semantic-rich datasets. This underscores its effectiveness while also enabling significantly faster record linkage computation.The integration of features that make the embeddings easy to use for SMPC (e.g., binary vectors), combined with robust performance and resource efficiency, establishes our framework as a highly valuable tool for real-world record linkage applications (e.g. in healthcare, financial services) including privacy-preserving record linkage.We make our framework’s current implementation available on GitHub1for ease of use and access.

People face with the huge amount of information on each day with the advent of big data era. The data amount stored and processed by Facebook, Twitter and other big social networks store (e.g. Instagram) is massive in those days. Online social networks provide great opportunity for propagation of almost any type of information. It's actually much much easier to disseminate an idea/knowledge than previous times. Naturally, this creates information validity and immediate curiosity about mass evaluation problem in general. In this regard, sentimental polarity detection in social media (e.g.Classification of a tweet as negative or positive or neutral) is highly valuable for certain institutions, organizations. The study's main focus is to classify negative, positive and neutral approaches of three (3) annotated twitter datasets. Effect of oversampling, unigram features and other features on overall and class-based accuracy ratios is worked on the datasets. Baseline is reached in dataset-2 experiments. 88% overall accuracy was observed in dataset-1 experiments which outperforms the prior art.Unigram features has shown significant effect on overall accuracy, class-based accuracy balance.