mausumi goswami - Academia.edu (original) (raw)

Related Authors

P. Sasikala

delowar sikder

MD AKIL RAIHAN IFTEE

José Bidarra

lalit kane

University of petroleum & energy studies

Wahab Khan

Sarang Shaikh

Uploads

Papers by mausumi goswami

Research paper thumbnail of Text Analysis and Speech to Sign Language Translation

American Sign Language is the most widely known sign language. There are more sign languages used... more American Sign Language is the most widely known sign language. There are more sign languages used as well. The current scenario is that deaf individuals have already learned the language and use it for their daily communication. The only hurdle is that normal person have to learn the sign language In this paper an architecture is proposed based on machine learning. The system is designed in 3 modules speech to text, text to sign and sign to animation. The first module is implemented using a speech recognition API. The second through a machine learning algorithm. The end module consists of a 3D animated avatar. .

Research paper thumbnail of Intelligent approach to automate a system for simulation of nanomaterials

ISET INTERNATIONAL CONFERENCE ON APPLIED SCIENCE & ENGINEERING (CASE 2021)

Nanomaterial composites are generally found to have great thermal properties and hence have witne... more Nanomaterial composites are generally found to have great thermal properties and hence have witnessed an increasing demand in the recent years for manufacturing of efficient miniature electronic devices. The process of finding the right composites that exhibit the desired properties is a rather tedious task involving a lot of trial and error in the current scenario. This paper proposes a methodology to digitize and automate this entire process by administering certain efficient practices of assessing the properties of nanomaterial like Coarse Grained Molecular Dynamics thus resulting in faster simulations.

Research paper thumbnail of An Empirical Analysis of Preprocessing Tasks for Unstructured Data

Journal of Computational and Theoretical Nanoscience, 2020

Nowadays computing systems are able to learn, reason, hear and see. Enormous amount of new opport... more Nowadays computing systems are able to learn, reason, hear and see. Enormous amount of new opportunities are created by artificial Intelligence. Artificial Intelligence has given two promising technologies. One such technology is NLP. Text Mining is also a promising area. These technologies enable and empower users to transform/map the key content in texts lying in documents into quantitative insight or to draw conclusion. Document Classification using supervised and unsupervised learning has a huge list of applications. Soft computing techniques like fuzzy logic helps to find a practical solution. This work proposes a methodology which is found to be significant while clustering unstructured data. The empirical analysis is included to demonstrate the improvement in document representation by using sparsification. Experiments are conducted on 5500 emails. Proposed methodology showed a significant improvement in representation of unstructured data.

Research paper thumbnail of CNNC: a common nearest neighbour clustering approach for gene expression data

International Journal of Computational Vision and Robotics, 2011

We present an effective common nearest neighbour-based clustering technique (CNNC) for finding cl... more We present an effective common nearest neighbour-based clustering technique (CNNC) for finding clusters over gene expression data. CNNC attempts to find all the clusters over gene expression data qualitatively. Our algorithm works by finding clusters using a nearest ...

Research paper thumbnail of Text Analysis and Speech to Sign Language Translation

American Sign Language is the most widely known sign language. There are more sign languages used... more American Sign Language is the most widely known sign language. There are more sign languages used as well. The current scenario is that deaf individuals have already learned the language and use it for their daily communication. The only hurdle is that normal person have to learn the sign language In this paper an architecture is proposed based on machine learning. The system is designed in 3 modules speech to text, text to sign and sign to animation. The first module is implemented using a speech recognition API. The second through a machine learning algorithm. The end module consists of a 3D animated avatar. .

Research paper thumbnail of Intelligent approach to automate a system for simulation of nanomaterials

ISET INTERNATIONAL CONFERENCE ON APPLIED SCIENCE & ENGINEERING (CASE 2021)

Nanomaterial composites are generally found to have great thermal properties and hence have witne... more Nanomaterial composites are generally found to have great thermal properties and hence have witnessed an increasing demand in the recent years for manufacturing of efficient miniature electronic devices. The process of finding the right composites that exhibit the desired properties is a rather tedious task involving a lot of trial and error in the current scenario. This paper proposes a methodology to digitize and automate this entire process by administering certain efficient practices of assessing the properties of nanomaterial like Coarse Grained Molecular Dynamics thus resulting in faster simulations.

Research paper thumbnail of An Empirical Analysis of Preprocessing Tasks for Unstructured Data

Journal of Computational and Theoretical Nanoscience, 2020

Nowadays computing systems are able to learn, reason, hear and see. Enormous amount of new opport... more Nowadays computing systems are able to learn, reason, hear and see. Enormous amount of new opportunities are created by artificial Intelligence. Artificial Intelligence has given two promising technologies. One such technology is NLP. Text Mining is also a promising area. These technologies enable and empower users to transform/map the key content in texts lying in documents into quantitative insight or to draw conclusion. Document Classification using supervised and unsupervised learning has a huge list of applications. Soft computing techniques like fuzzy logic helps to find a practical solution. This work proposes a methodology which is found to be significant while clustering unstructured data. The empirical analysis is included to demonstrate the improvement in document representation by using sparsification. Experiments are conducted on 5500 emails. Proposed methodology showed a significant improvement in representation of unstructured data.

Research paper thumbnail of CNNC: a common nearest neighbour clustering approach for gene expression data

International Journal of Computational Vision and Robotics, 2011

We present an effective common nearest neighbour-based clustering technique (CNNC) for finding cl... more We present an effective common nearest neighbour-based clustering technique (CNNC) for finding clusters over gene expression data. CNNC attempts to find all the clusters over gene expression data qualitatively. Our algorithm works by finding clusters using a nearest ...

Log In