Search

Speech enables clear and powerful idea transmission. The human voice, rich in tone and emotion, holds unique beauty and significance in daily life. Vocal pitches vary by gender and are influenced by emotions and languages. While people naturally perceive these nuances, machines often struggle to capture these subtle distinctions. Machines may struggle to detect these nuances, but people effortlessly perceive them. This project aims to use various machine learning (ML) and deep learning (DL) techniques to reliably determine an individual's gender from a corpus of Bengali conversations. Our dataset comprises 3185 Bengali speeches, with 1100 delivered by males, 1035 by women, and 1050 by those who identify as third gender. We employed six distinct feature extraction techniques to examine the audio data: roll-off, spectral centroid, chroma-stft, spectral bandwidth, zero crossing rate, and Mel-frequency cepstral coefficients (MFCC). Extreme gradient boosting (XGBoost), support vector machines (SVM), Knearest neighbors (KNN), decision trees classifier (DTC), and random forest (RF) were employed as the five ML algorithms to comprehensively analyze the dataset. For a full study, we also included 1D convolutional neural networks (CNN) from the DL area. The 1D CNN performed extraordinarily well, exceeding the accuracy of all other algorithms with a stunning 99.37%. [ABSTRACT FROM AUTHOR]

Nowadays addiction to drugs and alcohol has become a significant threat to the youth of the society as Bangladesh's population. So, being a conscientious member of society, we must go ahead to prevent these young minds from life-threatening addiction. In this paper, we approach a machine-learning-based way to forecast the risk of becoming addicted to drugs using machine-learning algorithms. First, we find some significant factors for addiction by talking to doctors, drug-addicted people, and read relevant articles and write-ups. Then we collect data from both addicted and nonaddicted people. After preprocessing the data set, we apply nine conspicuous machine learning algorithms, namely k-nearest neighbors, logistic regression, SVM, naïve bayes, classification, and regression trees, random forest, multilayer perception, adaptive boosting, and gradient boosting machine on our processed data set and measure the performances of each of these classifiers in terms of some prominent performance metrics. Logistic regression is found outperforming all other classifiers in terms of all metrics used by attaining an accuracy approaching 97.91%. On the contrary, CART shows poor results of an accuracy approaching 59.37% after applying principal component analysis. [ABSTRACT FROM AUTHOR]

N-gram based language models are very popular and extensively used statistical methods for solving various natural language processing problems including grammar checking. Smoothing is one of the most effective techniques used in building a language model to deal with data sparsity problem. Kneser-Ney is one of the most prominently used and successful smoothing technique for language modelling. In our previous work, we presented a Witten-Bell smoothing based language modelling technique for checking grammatical correctness of Bangla sentences which showed promising results outperforming previous methods. In this work, we proposed an improved method using Kneser-Ney smoothing based n-gram language model for grammar checking and performed a comparative performance analysis between Kneser-Ney and Witten-Bell smoothing techniques for the same purpose. We also provided an improved technique for calculating the optimum threshold which further enhanced the the results. Our experimental results show that, Kneser-Ney outperforms Witten-Bell as a smoothing technique when used with n-gram LMs for checking grammatical correctness of Bangla sentences. [ABSTRACT FROM AUTHOR]