Browsing by Author "Kowsher, Md."

Now showing 1 - 12 of 12

A Comprehensive Review on Big Data for Industries
(IEEE, 2022-12-26) Sarker, Supriya; Arefin, Mohammad Shamsul; Kowsher, Md.; Bhuiyan, Touhid; Kwon, Oh-Jin; Dhar, Pranab Kumar
Technological advancements in large industries like power, minerals, and manufacturing are generating massive data every second. Big data techniques have opened up numerous opportunities to utilize massive datasets in several effective ways to improve the efficacy of related industries. This paper presents a review of big data technologies used in the power, mineral, and manufacturing industries for various purposes. We analyze the meta-data of the collected papers before reviewing and selecting papers by applying selection criteria and paper quality assessment strategy. Then we propose a taxonomy of big data application areas in the power, mineral, and manufacturing industries. We have studied current big data architectures and techniques implemented in industry sectors and have uncovered the big data research gaps in industry sectors. To address the gaps, we point out some relevant research questions and, to answer the questions, we make some future research recommendations that might explore interesting research ideas for building a big data-driven industry. As the careful use of big data benefits every other industry sector; hence, supportive big data frameworks need to be developed to speed up the big data analysis process. Proper multi-dimensional big data assessment is also needed to take into account for serving effective data analysis tasks. Industry automation is also heavily influenced by the proper utilization of big data. While an intelligent agent can make many processes and heavy production loads in the manufacturing industry, it can work in a risky environment such as mines efficiently. To train agents for working in a specific environment big data can be used.
Bangla Intelligence Question Answering System Based on Mathematics and Statistics
(22nd International Conference on Computer and Information Technology, IEEE, 2019-12-20) Kowsher, Md.; Rahman, M M Mahabubur; Ahmed, Sk Shohorab; Prottasha, Nusrat Jahan
The Bangla Informative Question Answering System (BIQAS) is a significant Machine Learning (ML) technique that helps a user to trace relevant information by Bengali Natural Language Processing (BNLP). In this research paper, we have applied three mathematical and statistical procedures for BIQAS based on question answering data. These procedures are cosine similarity, Jaccard similarity, and Naive Bayes algorithm. The cosine similarity has interacted with dimension reduction technique SVD on user questions and questions answering data in order to reduce the space and time complexity. These procedures of this research are separated into two parts: pre-processing data and establishment of a relationship between user's questions and contained informative questions. We have got 93.22% accurate answer by using cosine similarity, 84.64% by Jaccard similarity and 91.31% by Naïve Bayes algorithm.
Doly
(1st International Conference on Advances in Science, Engineering and Robotics Technology 2019, ICASERT, IEEE, 2019-12-19) Kowsher, Md.; Tithi, Farhana Sharmin; Alam, F, M Ashraful; Huda, Mohammad Nurul; Moheuddin, Mir Md
This Scientific Research paper is a procedure of an automated system "Doly: Bengali Chabot" which gives a reply to a user query on behalf of a human for the education system in the Bengali language. This is an AI-based Chabot, mainly based on machine learning algorithms and Bengali Natural Language Processing (BNLP). The machine gets embedded with this knowledge to identify the desired sentences and making a decision within itself, as a response to answer questions. There are many English Chabot’s which used in education, web query, banking sector & various sectors. In this research, we have propounded a complete data-driven retrieval based closed domain Chabot which is easily colloquy in the Bengali language with the users. We've created the train function adapter to train the Doly by encoding (encoding="utf8") our corpus from bot data. An input adapter has been created to take input and for output, an output adapter has been created to generate automated responses to a user's input. We have also used a machine learning algorithm like search algorithm for finding an appropriate list of matching results from the corpus and use Naïve Bayesian algorithm to generate the right answer from data. The main aim of this Chabot based system is to bridge the gap between the knowledge sources by providing instant replies to the questions and queries that have to ask in the Bengali language.
Knowledge-base Optimization to Reduce the Response Time of Bangla Chatbot
(2020 Joint 9th International Conference on Informatics, Electronics & Vision (ICIEV) and 2020 4th International Conference on Imaging, Vision & Pattern Recognition (icIVPR), IEEE, 2020-01-07) Kowsher, Md.; Tahabilder, Anik; Sanjid, Md. Zahidul Islam; Prottasha, Nusrat Jahan; Sarker, Md. Murad Hossain
Chatbots have been very popular in recent years for being able to serve as a customer representative, a language learner and so forth. Long short-term memory abbreviated as LSTM is a ubiquitous artificial recurrent neural network that is frequently being used for the chatbot. Nevertheless, if a user makes the line break of sequence, then it is rare to inform the right information without the impact of the previous sequence. As a result, in case of a help desk chatbot, LSTM is not the best option for taking steps of the right information. On the other hand, mathematical and statistical procedures are prominently useful for providing the proper knowledge without having back the impact of sequence. Still, it takes more execution time to respond. The goal of this paper is to present the optimal chatbot for the lowest execution time and three mathematical and statistical strategies for Bangla Intelligence chatbot in light of information obtained from Noakhali Science and Technology University (NSTU). As the procedures, we have followed cosine similarity, Jaccard similarity, and Naive Bayes classifier. To reduce the response time, we decorated the whole path into a 3-depth tree, such as a question, topic, and answer. We have compared the performance of the selected strategies where the best accuracy was 93.22% using the cosine similarity. Contribution-This paper presents techniques to reduce the response time of statistical and mathematical Bangla Chatbot.
Lemmatization Algorithm Development for Bangla Natural Language Processing
(2020 Joint 9th International Conference on Informatics, Electronics & Vision (ICIEV) and 2020 4th International Conference on Imaging, Vision & Pattern Recognition (icIVPR), IEEE, 2020-01-07) Kowsher, Md.; Tahabilder, Anik; Sarker, Md Murad Hossain; Sanjid, Md. Zahidul Islam; Prottasha, Nusrat Jahan
Natural language processing (NLP) finds enormous applications in autonomous communication, while lemmatization is an essential preprocessing technique for simplification of a word to its origin-word in NLP. However, there is scarcity of effective algorithms in Bangla NLP. This leads us to develop a useful Bangla language lemmatization tool. Usually, some rule base stemming processes play the vital role of lemmatization in Bangla language processing as there is lack of Bangla lemmatization tool. In this paper, we propose a Bangla lemmatization framework using three effective lemmatization techniques based on data structures and dynamic programming. We have used Trie algorithm and developed a mapping algorithm named “Dictionary Based Search by Removing Affix (DBSRA)” based on data structure. We have applied both Trie and DBSRA lemmatization and selected the better one by considering the Levenshtein distance between the lemma and the original word. Eventually, we have experimented with Bangla language lemmatization among all three techniques and the framework. Among the three proposed techniques, the DBSRA performed better compared to others with an accuracy of 93.1 percent. The framework, developed by fusing three algorithms, came out with the highest efficiency of 95.89 percent. Contribution-This paper presents the development of three lemmatization algorithms and their fusion to develop a framework for Bangla Natural Language Processing.
LSTM-ANN & BiLSTM-ANN
(Scopus, 2021) Kowsher, Md.; Tahabilder, Anik; Sanjid, Md. Zahidul Islam; Prottasha, Nusrat Jahan; Uddin, Md. Shihab; Hossain, Md Arman; Jilani, Md. Abdul Kader
Machine learning is getting more and more advanced with the progression of state-of-the-art technologies. Since existing algorithms do not provide a palatable learning performance most often, it is necessary to carry on the trail of upgrading the current algorithms incessantly. The hybridization of two or more algorithms can potentially increase the performance of the blueprinted model. Although LSTM and BiLSTM are two excellent far and widely used algorithms in natural language processing, there still could be room for improvement in terms of accuracy via the hybridization method. Thus, the advantages of both RNN and ANN algorithms can be obtained simultaneously. This paper has illustrated the deep integration of BiLSTM-ANN (Fully Connected Neural Network) and LSTM-ANN and manifested how these integration methods are performing better than single BiLSTM, LSTM and ANN models. Undertaking Bangla content classification is challenging because of its equivocalness, intricacy, diversity, and shortage of relevant data, therefore, we have executed the whole integrated models on the Bangla content classification dataset from newspaper articles. The proposed hybrid BiLSTM-ANN model beats all the implemented models with the most noteworthy accuracy score of 93% for both validation & testing. Moreover, we have analyzed and compared the performance of the models based on the most relevant parameters.
Machine Learning Based Recommendation Systems for the Mode of Childbirth
(Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, Springer, 2020-07-30) Kowsher, Md.; Prottasha, Nusrat Jahan; Tahabilder, Anik; Islam, Md. Babul
Machine learning method gives a learning technique that can be applied to extract information from data. Lots of researches are being conducted that involves machine learning techniques for medical diagnosis, prediction and treatment. The goal of this study is to perform several machine learning actions for finding the appropriate mode of birth (cesarean or normal) to minimize maternal mortality rate. To generate a computer-aided decision for selecting between the most common way of baby birth, C-section and vaginal birth, we have used supervised machine learning to train our classification model. A dataset consists of the information of 13,527 delivery patients has been collected from Tarail Upazilla Health complex, Bangladesh. We have implemented nine machine learning classifier algorithms over the whole datasets and compared the performances of all those proposed techniques. The computer recommended mode of baby delivery suggested by the most convincing method named “impact learning,” showed an accuracy of 0.89089172 with the F1 value of 0.877871741.
Predicting the Appropriate Mode of Childbirth using Machine Learning Algorithm
(Scopus, 2021) Kowsher, Md.; Tahabilder, Anik; Prottasha, Nusrat Jahan; Rakib, Md. Abdur-; Alam, Md. Shameem; Habib, Kaiser
—A woman's satisfaction with childbirth may have immediate and long-term effects on her health as well as on the relationship with her newborn child. The mode of baby delivery is genuinely vital to a delivery patient and her infant child. It might be a crucial factor for ensuring the safety of both the mother and the child. During the baby delivery, decision-making within a short time becomes very challenging for the physician. Besides, humans may make wrong decisions selecting the appropriate delivery mode of childbirth. A wrong decision increases the mother's life risk and can also be harmful to the newborn baby's health. Computer-aided decision-making can be an excellent solution to this problem. Considering this scope, we have built a supervised machine learning-based decision-making model to predict the most suitable childbirth mode that will reduce this risk. This work has applied 32 supervised classifier algorithms and 11 training methods on the real childbirth dataset from the Tarail Upazilla Health complex, Kishorganj, Bangladesh. We have also analyzed the result and compared them using various statistical parameters to determine the best-performed model. The quadratic discriminant analysis has shown the highest accuracy of 0.979992 with the F1 score of 0.979962. Using this model to decide the appropriate labor mode may significantly reduce maternal and infant health risks.
SDSV
(Scopus, 2021) Kowsher, Md.; Hossen, Imran; Tahabilder, Anik; Prottasha, Nusrat Jahan; Sarker, Md. Murad Hossain; Ahasan, Nazmul; Hoque, Md Imdadul
Data classification is one of the most fundamental tasks that can be accomplished by supervised machine learning. There exists a lot of algorithms, and they have the specific case of uses. Different classification methods follow different techniques to map the relationship between input and output. This article proposes an angle measurement-based classification technique called Support Directional Shifting Vectors (SDSV) to segment a spectral domain into regions with a very effective solution for classification problems. This method introduces two shifting vectors, named Support Direction Vector (SDV) and Support Origin Vector (SOV). These vectors are formed as a linear function to measure cosine-angle by the dot product of two separated data classes, named target data points and non-target data points. Considering the target class samples, the vectors get aligned in a way that the angle with the target class gets minimized, while the angle with the non-target class gets maximized. The error in the position of the linear function has been modeled as the loss function. Then, the vector position is updated iteratively by optimizing this loss function using a gradient descent algorithm. We have used this model to classify data from two different machine learning datasets to evaluate the performance of the proposed method. Finally, we have carefully examined the results and compared them with the other standard classification algorithms. In summary, the proposed SDSV algorithm has shown a comparable accuracy compared to different standard algorithms.
Short-Term Rainfall Prediction Using Supervised Machine Learning
(Elsevier, 2023-01-15) Prottasha, Nusrat Jahan; Tahabilder, Anik; Kowsher, Md.; Mia, Md Shanon; Kobra, Khadiza Tul
Floods and rain significantly impact the economy of many agricultural countries in the world. Early prediction of rain and floods can dramatically help prevent natural disaster damage. This paper presents a machine learning and data-driven method that can accurately predict short-term rainfall. Various machine learning classification algorithms have been implemented on an Australian weather dataset to train and develop an accurate and reliable model. To choose the best suitable prediction model, diverse machine learning algorithms have been applied for classification as well. Eventually, the performance of the models has been compared based on standard performance measurement metrics. The finding shows that the hist gradient boosting classifier has given the highest accuracy of 91%, with a good F1 value and receiver operating characteristic, the area under the curve score.
Support Directional Shifting Vector
(Emerging Science Journal, 2021) Kowsher, Md.; Hossen, Imran; Tahabilder, Anik; Prottasha, Nusrat Jahan; Habib, Kaiser; Azmi, Zafril Rizal M.
Machine learning models have been very popular nowadays for providing rigorous solutions to complicated real-life problems. There are three main domains named supervised, unsupervised, and reinforcement. Supervised learning mainly deals with regression and classification. There exist several types of classification algorithms, and these are based on various bases. The classification performance varies based on the dataset velocity and the algorithm selection. In this article, we have focused on developing a model of angular nature that performs supervised classification. Here, we have used two shifting vectors named Support Direction Vector (SDV) and Support Origin Vector (SOV) to form a linear function. These vectors form a linear function to measure cosine-angle with both the target class data and the non-target class data. Considering target data points, the linear function takes such a position that minimizes its angle with target class data and maximizes its angle with non-target class data. The positional error of the linear function has been modelled as a loss function which is iteratively optimized using the gradient descent algorithm. In order to justify the acceptability of this method, we have implemented this model on three different standard datasets. The model showed comparable accuracy with the existing standard supervised classification algorithm.
Transfer Learning for Sentiment Analysis Using BERT Based Supervised Fine-Tuning
(Daffodil International University, 2022-06-13) Prottasha, Nusrat Jahan; As Sami, Abdullah; Kowsher, Md.; Murad, Saydul Akbar; Bairagi, Anupam Kumar; Masud, Mehedi; Baz, Mohammed
The growth of the Internet has expanded the amount of data expressed by users across multiple platforms. The availability of these different worldviews and individuals’ emotions empowers sentiment analysis. However, sentiment analysis becomes even more challenging due to a scarcity of standardized labeled data in the Bangla NLP domain. The majority of the existing Bangla research has relied on models of deep learning that significantly focus on context-independent word embeddings, such as Word2Vec, GloVe, and fastText, in which each word has a fixed representation irrespective of its context. Meanwhile, context-based pre-trained language models such as BERT have recently revolutionized the state of natural language processing. In this work, we utilized BERT’s transfer learning ability to a deep integrated model CNN-BiLSTM for enhanced performance of decision-making in sentiment analysis. In addition, we also introduced the ability of transfer learning to classical machine learning algorithms for the performance comparison of CNN-BiLSTM. Additionally, we explore various word embedding techniques, such as Word2Vec, GloVe, and fastText, and compare their performance to the BERT transfer learning strategy. As a result, we have shown a state-of-the-art binary classification performance for Bangla sentiment analysis that significantly outperforms all embedding and algorithms.