Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230910
Sujit Basu, M. Hossen, M. Hanawa
Online polling-based passive optical network (PON) effectively reduces the ideal time between two successive time cycles and large grant delay for optical network units (ONUs) than the offline polling-based PON. However, it introduced some other draw backs like unfair bandwidth allocation to the ONUs in different traffic load conditions and absent of priority scheduling to early serve the mostly busy ONUs. In this paper, we propose a new dynamic data transmission sequence algorithm for cluster-based PON that accommodate the advantages of both the existing offline and online polling-based schemes. In the proposed cluster-based PON, in each time cycle, optical line terminal divides all the active ONUs in a number of clusters and unused bandwidth of a cluster is added as a surplus bandwidth to the next cluster. Computer simulated results is obtained in the case of three different performance parameters where the proposed scheme outperforms than the existing offline and online poling-based schemes.
{"title":"A New Polling Algorithm for Dynamic Data Transmission Sequence of Cluster-based PON","authors":"Sujit Basu, M. Hossen, M. Hanawa","doi":"10.1109/TENSYMP50017.2020.9230910","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230910","url":null,"abstract":"Online polling-based passive optical network (PON) effectively reduces the ideal time between two successive time cycles and large grant delay for optical network units (ONUs) than the offline polling-based PON. However, it introduced some other draw backs like unfair bandwidth allocation to the ONUs in different traffic load conditions and absent of priority scheduling to early serve the mostly busy ONUs. In this paper, we propose a new dynamic data transmission sequence algorithm for cluster-based PON that accommodate the advantages of both the existing offline and online polling-based schemes. In the proposed cluster-based PON, in each time cycle, optical line terminal divides all the active ONUs in a number of clusters and unused bandwidth of a cluster is added as a surplus bandwidth to the next cluster. Computer simulated results is obtained in the case of three different performance parameters where the proposed scheme outperforms than the existing offline and online poling-based schemes.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"22 1","pages":"271-274"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77198302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9231034
Md. Raffael Maruf, Md. Omar Faruque, Salman Mahmood, Nazmun Nahar Nelima, Md. Golam Muhtasim, Md.Jahedul Alam Pervez
Though Bangla Automatic Speech Recognition (ASR) started its journey since a long time ago, a paltry amount of work is done on Convolutional Neural Network (CNN) based ASR. In this paper, we propose an ASR made with CNN where the performance of two feature extraction methods, namely Mel Frequency Cepstral Coefficients (MFCC) and Relative Spectral Transform - Perceptual Linear Prediction (RASTA-PLP) are compared on Bangla isolated words consisting of digits and speech commands. This paper contributes to the literature of Bangla ASR in three ways. Firstly, Effects of noise is experimented on Bangla speech commands as well as isolated words in CNN based ASR. Secondly, the performance of MFCC and RASTA-PLP are compared in noisy environment using CNN based classifier. Lastly, state-of-the-art accuracy is achieved in CNN based ASR which is 93.18%.
{"title":"Effects of Noise on RASTA-PLP and MFCC based Bangla ASR Using CNN","authors":"Md. Raffael Maruf, Md. Omar Faruque, Salman Mahmood, Nazmun Nahar Nelima, Md. Golam Muhtasim, Md.Jahedul Alam Pervez","doi":"10.1109/TENSYMP50017.2020.9231034","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9231034","url":null,"abstract":"Though Bangla Automatic Speech Recognition (ASR) started its journey since a long time ago, a paltry amount of work is done on Convolutional Neural Network (CNN) based ASR. In this paper, we propose an ASR made with CNN where the performance of two feature extraction methods, namely Mel Frequency Cepstral Coefficients (MFCC) and Relative Spectral Transform - Perceptual Linear Prediction (RASTA-PLP) are compared on Bangla isolated words consisting of digits and speech commands. This paper contributes to the literature of Bangla ASR in three ways. Firstly, Effects of noise is experimented on Bangla speech commands as well as isolated words in CNN based ASR. Secondly, the performance of MFCC and RASTA-PLP are compared in noisy environment using CNN based classifier. Lastly, state-of-the-art accuracy is achieved in CNN based ASR which is 93.18%.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"40 1","pages":"1564-1567"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82273913","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230931
F. Alam, M. Khan
In this article a graphene based direct-fed multiband planar internal antenna (PIA) for wireless communication is proposed. Graphene, one of the most disruptive materials, is used to enhance the performance of this antenna. Graphene based antennas present noteworthy perfections for the maximum parameters than copper based antennas. Designing and simulating in CST over 0 to 3 THz, it is observed that the PIA indicates the highest efficiency over the different types of graphene based patch antennas. It performs with almost 98 % efficiency.
{"title":"Design an Efficient Graphene Based Antenna for Wireless Applications","authors":"F. Alam, M. Khan","doi":"10.1109/TENSYMP50017.2020.9230931","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230931","url":null,"abstract":"In this article a graphene based direct-fed multiband planar internal antenna (PIA) for wireless communication is proposed. Graphene, one of the most disruptive materials, is used to enhance the performance of this antenna. Graphene based antennas present noteworthy perfections for the maximum parameters than copper based antennas. Designing and simulating in CST over 0 to 3 THz, it is observed that the PIA indicates the highest efficiency over the different types of graphene based patch antennas. It performs with almost 98 % efficiency.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"15 1","pages":"1460-1463"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81392609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230751
Sujanya Kumari. T, L. P. Roy
This paper presents radar signal processing techniques for measuring frequency of vibrating diaphragm through-the-wall, using a HB100 Doppler radar. The micro-Doppler features of radar are useful in determining the frequency of sound waves. Therefore, Doppler features are apprehended in a signal model for analysing through-the-wall diaphragm vibration. The same model is modified for Multiple Signal Classification (MUSIC) algorithm, in accurate measurement of frequency of sound waves, as conventional Fast Fourier Transform(FFT) algorithm incurs undesired additional frequency components. The experimental results presented in this paper signify that the background subtraction technique is preferred for through-the-wall scenario as it eliminates undesired frequency from the received signal and thus, measures the Doppler frequency precisely. Therefore, HB100 radar can be used for measuring through-the-wall vibration of sound source from its micro-motion signature.
{"title":"Through-the-Wall HB100 Radar Signal Processing for Estimating Frequency of Vibrating Diaphragm","authors":"Sujanya Kumari. T, L. P. Roy","doi":"10.1109/TENSYMP50017.2020.9230751","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230751","url":null,"abstract":"This paper presents radar signal processing techniques for measuring frequency of vibrating diaphragm through-the-wall, using a HB100 Doppler radar. The micro-Doppler features of radar are useful in determining the frequency of sound waves. Therefore, Doppler features are apprehended in a signal model for analysing through-the-wall diaphragm vibration. The same model is modified for Multiple Signal Classification (MUSIC) algorithm, in accurate measurement of frequency of sound waves, as conventional Fast Fourier Transform(FFT) algorithm incurs undesired additional frequency components. The experimental results presented in this paper signify that the background subtraction technique is preferred for through-the-wall scenario as it eliminates undesired frequency from the received signal and thus, measures the Doppler frequency precisely. Therefore, HB100 radar can be used for measuring through-the-wall vibration of sound source from its micro-motion signature.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"343 1","pages":"851-854"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76389854","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230852
T. Chakrabarty, R. Saha, M. Faysal, M. R. Bishal, M. S. Hossain
The performance of beamformer deteriorates due to signal mismatch caused by look direction disparity and scattering. To overcome this issue, a modified robust Linearly Constrained Minimum Variance(LCMV) beamforming method using variable optimal diagonal loading is addressed in this paper. It uses null constraints to reject interference signals and diagonal loading to increase the robustness of the beamformer against any steering vector mismatch. The performance of the proposed method has been observed for several steering vector mismatch condition. In every scenario, it achieves higher output signal to interference plus noise ratio (SINR) at lower snapshots than the existed beamforming method. Simulations are performed in MATLAB for finding out the performance of the proposed method.
{"title":"Performance Investigation of Robust Linearly Constrained Minimum Variance Beamforming for Uniform Circular Array","authors":"T. Chakrabarty, R. Saha, M. Faysal, M. R. Bishal, M. S. Hossain","doi":"10.1109/TENSYMP50017.2020.9230852","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230852","url":null,"abstract":"The performance of beamformer deteriorates due to signal mismatch caused by look direction disparity and scattering. To overcome this issue, a modified robust Linearly Constrained Minimum Variance(LCMV) beamforming method using variable optimal diagonal loading is addressed in this paper. It uses null constraints to reject interference signals and diagonal loading to increase the robustness of the beamformer against any steering vector mismatch. The performance of the proposed method has been observed for several steering vector mismatch condition. In every scenario, it achieves higher output signal to interference plus noise ratio (SINR) at lower snapshots than the existed beamforming method. Simulations are performed in MATLAB for finding out the performance of the proposed method.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"46 1","pages":"1201-1204"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76000409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230462
Rifat Hasan, M. Mowla, Nabila Hoque
The next major step in mobile communications technology beyond the current LTE-A is 5G wireless technologies. With the rapid progression of 5G, massive MIMO innovation has turned out to be one of the most significant advances in 5G networks by facilitating beamforming technology. In this paper, we have simulated and analyzed the statistical propagation channel model for 5G massive MIMO millimeter-wave (mmWave) cellular communication in the ultrawideband unlicensed spectrum of 60 GHz. This investigation is based on IR (impulse response) channel model named NYUSIM which is a geometrical based channel modeling tool developed by NYUWireless. We have used the drop based mode to determine the propagation channel coefficients. We investigated the network performance by considering the directional, omnidirectional power delay profiles (PDPs) and path loss. Analyzing the propagation channel is fundamental to define the actions of the channel response of the wireless stream link in the specific environment to reach the 5G expectations.
{"title":"Performance Estimation of Massive MIMO Drop-based Propagation Channel Model for mmWave Communication","authors":"Rifat Hasan, M. Mowla, Nabila Hoque","doi":"10.1109/TENSYMP50017.2020.9230462","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230462","url":null,"abstract":"The next major step in mobile communications technology beyond the current LTE-A is 5G wireless technologies. With the rapid progression of 5G, massive MIMO innovation has turned out to be one of the most significant advances in 5G networks by facilitating beamforming technology. In this paper, we have simulated and analyzed the statistical propagation channel model for 5G massive MIMO millimeter-wave (mmWave) cellular communication in the ultrawideband unlicensed spectrum of 60 GHz. This investigation is based on IR (impulse response) channel model named NYUSIM which is a geometrical based channel modeling tool developed by NYUWireless. We have used the drop based mode to determine the propagation channel coefficients. We investigated the network performance by considering the directional, omnidirectional power delay profiles (PDPs) and path loss. Analyzing the propagation channel is fundamental to define the actions of the channel response of the wireless stream link in the specific environment to reach the 5G expectations.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"108 1","pages":"461-464"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87618684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230867
Joydhriti Choudhury, Faisal Bin Ashraf, Arif Shakil, Nahian Raonak
Ethnic cleansing of Rohingya ethnicity from the Rakhine state of Myanmar has made life miserable for more than half million persons who had fled away with their life from their own country. They have taken shelter and and have been living in in the resource-poor side of Bangladesh. Immense size of refugee population makes it challenging to accommodate all the needs. In case of refugee rehabilitation, all the refugees are given shelter in small camps. Different camps have different types of people and needs. However, not all the needs can be met altogether. So, prioritizing needs will make the rehabilitation process more effective. In this paper, we have used machine learning techniques to identify an effective model which predicts the needs based on priority. This learned model can be used to predict the prioritized needs for different camps while rehabilitation process goes on. Our experiments disclosed that Random Forest ensemble methods work effectively.
{"title":"Predicting priority needs for Rehabilitation of refugees based on machine learning techniques from monitoring data of Rohingya refugees in Bangladesh","authors":"Joydhriti Choudhury, Faisal Bin Ashraf, Arif Shakil, Nahian Raonak","doi":"10.1109/TENSYMP50017.2020.9230867","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230867","url":null,"abstract":"Ethnic cleansing of Rohingya ethnicity from the Rakhine state of Myanmar has made life miserable for more than half million persons who had fled away with their life from their own country. They have taken shelter and and have been living in in the resource-poor side of Bangladesh. Immense size of refugee population makes it challenging to accommodate all the needs. In case of refugee rehabilitation, all the refugees are given shelter in small camps. Different camps have different types of people and needs. However, not all the needs can be met altogether. So, prioritizing needs will make the rehabilitation process more effective. In this paper, we have used machine learning techniques to identify an effective model which predicts the needs based on priority. This learned model can be used to predict the prioritized needs for different camps while rehabilitation process goes on. Our experiments disclosed that Random Forest ensemble methods work effectively.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"28 1","pages":"210-213"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87988271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230824
Md. Tanvir Emrose, A. Chowdhury, S. M. Mehedi Hasan
This paper presents a detailed model of 400 kV AC double circuit transmission line between Ashuganj and Bhulta substations of Bangladesh power system network. Frequency dependent model is for overhead transmission lines, constant parameter distributed line model for tower, variable tower footing resistance model, and IEEE standard model for surge arrester are used for the purpose. The complete model is tested and verified using actual data by employing lightning strike.
{"title":"Modeling of a 400 kV Transmission Line of Bangladesh Power System Network","authors":"Md. Tanvir Emrose, A. Chowdhury, S. M. Mehedi Hasan","doi":"10.1109/TENSYMP50017.2020.9230824","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230824","url":null,"abstract":"This paper presents a detailed model of 400 kV AC double circuit transmission line between Ashuganj and Bhulta substations of Bangladesh power system network. Frequency dependent model is for overhead transmission lines, constant parameter distributed line model for tower, variable tower footing resistance model, and IEEE standard model for surge arrester are used for the purpose. The complete model is tested and verified using actual data by employing lightning strike.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"78 1","pages":"1134-1139"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86193341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The significance of text summarization in the Natural Language Processing (NLP) community has now expanded because of the staggering increase in virtual textual materials. Text summary is the process created from one or multiple texts which convey important insight in a little form of the main text. Multiple text summarization technique assists to pick indispensable points of the original texts reducing time and effort require reading the whole document. The question was approached from a different point of view, in a different domain by using different concepts. Extractive and abstractive are the two main methods of summing up text. Though extractive summary is primarily concerned with what summary content the frequency of words, phrases, and sentences from the original document should be used. This research proposes a sentence based clustering algorithm (K-Means) for a single document. For feature extraction, we have used Gensim word2vec which is intended to automatically extract semantic topics from documents in the most efficient way possible.
{"title":"Automatic Text Summarization Using Gensim Word2Vec and K-Means Clustering Algorithm","authors":"Mofiz Mojib Haider, Md. Farhad Hossin, Hasibur Rashid Mahi, Hossain Arif","doi":"10.1109/TENSYMP50017.2020.9230670","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230670","url":null,"abstract":"The significance of text summarization in the Natural Language Processing (NLP) community has now expanded because of the staggering increase in virtual textual materials. Text summary is the process created from one or multiple texts which convey important insight in a little form of the main text. Multiple text summarization technique assists to pick indispensable points of the original texts reducing time and effort require reading the whole document. The question was approached from a different point of view, in a different domain by using different concepts. Extractive and abstractive are the two main methods of summing up text. Though extractive summary is primarily concerned with what summary content the frequency of words, phrases, and sentences from the original document should be used. This research proposes a sentence based clustering algorithm (K-Means) for a single document. For feature extraction, we have used Gensim word2vec which is intended to automatically extract semantic topics from documents in the most efficient way possible.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"6 1","pages":"283-286"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88885316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-06-05DOI: 10.1109/TENSYMP50017.2020.9230981
Ruhul Amin, Nabila Sabrin Sworna, Nahid Hossain
Text mining is the procedure of exploring large unorganized text data. Due to the availability of numerous amounts of text data through online blogs, newspapers and other media, text classification and categorization is the hot topic nowadays. Many researches have been done on this topic on English and other western languages. However, very few notable researches have been on Bangla language. Unavailability of a notable dataset in Bangla language is another burden to develop a highperformance text classification tool. In this paper, we have presented a Bangla news tags classification approach. The classification has been done entirely based on news titles only with parallel Convolutional Neural Network (CNN) which is a category of deep neural networks utilizing word-level data augmentation approach. Due to the unavailability of a proper and updated dataset on Bangla news titles and tags, we have developed our own dataset which consists of 88,968 news titles and tags by scrapping online newspapers. According to the classification result, our approach shows an accuracy of 93.47% which is the highest amongst the similar works.
{"title":"Multiclass Classification for Bangla News Tags with Parallel CNN Using Word Level Data Augmentation","authors":"Ruhul Amin, Nabila Sabrin Sworna, Nahid Hossain","doi":"10.1109/TENSYMP50017.2020.9230981","DOIUrl":"https://doi.org/10.1109/TENSYMP50017.2020.9230981","url":null,"abstract":"Text mining is the procedure of exploring large unorganized text data. Due to the availability of numerous amounts of text data through online blogs, newspapers and other media, text classification and categorization is the hot topic nowadays. Many researches have been done on this topic on English and other western languages. However, very few notable researches have been on Bangla language. Unavailability of a notable dataset in Bangla language is another burden to develop a highperformance text classification tool. In this paper, we have presented a Bangla news tags classification approach. The classification has been done entirely based on news titles only with parallel Convolutional Neural Network (CNN) which is a category of deep neural networks utilizing word-level data augmentation approach. Due to the unavailability of a proper and updated dataset on Bangla news titles and tags, we have developed our own dataset which consists of 88,968 news titles and tags by scrapping online newspapers. According to the classification result, our approach shows an accuracy of 93.47% which is the highest amongst the similar works.","PeriodicalId":6721,"journal":{"name":"2020 IEEE Region 10 Symposium (TENSYMP)","volume":"142 1","pages":"174-177"},"PeriodicalIF":0.0,"publicationDate":"2020-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88992163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：