With the advent of Deep Learning-based Artificial Neural Network models, Natural Language Processing (NLP) has witnessed significant improvements in textual data processing in terms of its efficiency and accuracy. However, the research is mostly restricted to high-resource languages such as English, and low-resource languages still suffer from a lack of available resources in terms of training datasets as well as models with even baseline evaluation results. Considering the limited availability of resources for low-resource languages, we propose a methodology for adapting self-attentive transformer-based architecture models (mBERT, mT5) for low-resource summarization, supplemented by the construction of a new baseline dataset (76.5k article, summary pairs) in a low-resource language, Urdu. Choosing news (a publicly available source) as the application domain has the potential to make the proposed methodology useful for reproducing in other languages with limited resources. Our adapted summarization model urT5 with up to 44.78% reduction in size as compared to mT5 can capture contextual information of the low-resource language effectively with an evaluation score (up to 46.35 ROUGE-1, 77 BERTScore) on par with state-of-the-art models in the high-resource language of English (PEGASUS: 47.21, BART: 45.14 on XSUM Dataset) . The proposed method provided a baseline approach toward extractive as well as abstractive summarization with competitive evaluation results in a limited resource setup.

Continuous proliferation of hate speech in different languages on social media has drawn significant attention from researchers in the past decade. Detecting hate speech is indispensable irrespective of the scale of use of language, as it inflicts huge harm on society. This work presents a first resource for classifying the severity of hate speech in addition to classifying offensive and hate speech content. Current research mostly limits hate speech classification to its primary categories, such as racism, sexism, and hatred of religions. However, hate speech targeted at different protected characteristics also manifests in different forms and intensities. It is important to understand varying severity levels of hate speech so that the most harmful cases of hate speech may be identified and dealt with earlier than the less harmful ones. In this work, we focus on detecting offensive speech, hate speech, and multiple levels of hate speech in the Urdu language. We investigate three primary target categories of hate speech: religion, racism, and national origin. We further divide these categories into levels based on the severity of hate conveyed. The severity levels are referred to as symbolization , insult , and attribution . A corpus comprising more than 20,000 tweets against the corresponding hate speech categories and severity levels is collected and annotated. A comprehensive experimentation scheme is applied using traditional as well as deep learning–based models to examine their impact on hate speech detection. The highest macro-averaged F-score yielded for detecting offensive speech is 86% while the highest F-scores for detecting hate speech with respect to ethnicity, national origin, and religious affiliation are 80%, 81%, and 72%, respectively. This shows that results are very encouraging and would provide a lead towards further investigation in this domain.

Increased connectivity has contributed greatly in facilitating rapid access to information and reliable communication. However, the uncontrolled information dissemination has also resulted in the spread of fake news. Fake news might be spread by a group of people or organizations to serve ulterior motives such as political or financial gains or to damage a country’s public image. Given the importance of timely detection of fake news, the research area has intrigued researchers from all over the world. Most of the work for detecting fake news focuses on the English language. However, automated detection of fake news is important irrespective of the language used for spreading false information. Recognizing the importance of boosting research on fake news detection for low resource languages, this work proposes a novel semantically enriched technique to effectively detect fake news in Urdu—a low resource language. A model based on deep contextual semantics learned from the convolutional neural network is proposed. The features learned from the convolutional neural network are combined with other n-gram-based features and are fed to a conventional majority voting ensemble classifier fitted with three base learners: Adaptive Boosting, Gradient Boosting, and Multi-Layer Perceptron. Experiments are performed with different models, and results show that enriching the traditional ensemble learner with deep contextual semantics along with other standard features shows the best results and outperforms the state-of-the-art Urdu fake news detection model.

The outstanding usage of web applications across the globe has enabled people to access desired information online with a few clicks. This has also enabled skilled attackers to compromise the availability, integrity and confidentiality of the data and information available on these websites. This paper proposes a framework for detecting and obstructing a large number of web attacks and scanning probes based on features of an HTTP (Hyper Text Transfer Protocol) request packet and also caters for POST HTTP data. We first trained four traditional machine learning models i.e. Decision Tree, Support Vector Machine (SVM), Naive Bayesian and Linear Regression by using a well-known publicly available dataset. It was found out that Decision Tree outperforms the rest in terms of performance and accuracy. Finally, a Convolutional Neural Network (CNN) based deep learning approach was implemented and tested on a well-known publicly available dataset. It results in optimal performance and an accuracy of 99.94%. The deep learning approach was enhanced by the introduction of a User Categorization Feature which uses cookies to categorise malicious attackers.

Recommender systems improve browsing experience of users for large amount of items by assisting selection and classification of items utilizing item metadata. The performance of recommender system usually deteriorates when implicit data is used with limited user interaction history also regarded as cold start (CS) problem. This paper proposes a model to address cold start problem using content based technique where user or item metadata is used to break this ice barrier. The proposed method utilizes the feature extraction techniques (such as term frequencyInverse document frequency(TF-IDF)) and word embedding technique (Word2Vec). These content features are then used to predict the ratings for CS items by constructing user profiles using stacked auto-encoder. Experiments performed on largest real world dataset provided by Movielens 20M shows that proposed model outperforms the state-of-the-art approaches in CS item scenario.

Recommender systems are increasingly being used in e-commerce websites to solve the problem of finding right kind of information. Collaborative filtering is considered as most promising method for recommendation because it recommends items based on common interests of users. Trust Aware Recommender Systems (TARS) is an enhancement of traditional recommendation systems to improve recommendation quality which uses trusted users for recommending an item to an active user. From literature, it is proven that including all trusted users in recommendation process reduces its performance so this research work performs a filtration process on users for reduction of trusted neighborhood of an active user. The main idea of this research work is to keep only those users in trusted neighborhood whose rating behavior is similar to an active user. Subspace clustering method is used for filtration process. The proposed algorithm uses both implicit and explicit trust for trust value calculation. The results demonstrates that the proposed algorithm improves results in terms of Mean Absolute Error and Coverage as compared to other conventional methods.

Traditional anti-virus technologies have failed to keep pace with proliferation of malware due to slow process of their signatures and heuristics updates. Similarly, there are limitations of time and resources in order to perform manual analysis on each malware. There is a need to learn from this vast quantity of data, containing cyber attack pattern, in an automated manner to proactively adapt to ever-evolving threats. Machine learning offers unique advantages to learn from past cyber attacks to handle future cyber threats. The purpose of this research is to propose a framework for multi-classification of malware into well-known categories by applying different machine learning models over corpus of malware analysis reports. These reports are generated through an open source malware sandbox in an automated manner. We applied extensive pre-modeling techniques for data cleaning, features exploration and features engineering to prepare training and test datasets. Best possible hyper-parameters are selected to build machine learning models. These prepared datasets are then used to train the machine learning classifiers and to compare their prediction accuracy. Finally, these results are validated through a comprehensive 10-fold cross-validation methodology. The best results are achieved through Gaussian Naive Bayes classifier with random accuracy of 96% and 10-Fold Cross Validation accuracy of 91.2%. The said framework can be deployed in an operational environment to learn from malware attacks for proactively adapting matching counter measures.