Pub Date : 2022-08-31DOI: 10.1109/CSI54720.2022.9924139
Bembamba Fulbert, O. T. Frédéric, Malo Sadouanouan, Yougbare Bernadette, O. Dominique
Uncontrolled crossbreeding between zebus and taurine cattle is jeopardizing the genetic heritage of West African taurines and their specific ability to resist trypanosomosis. to achieve any successful conservation policy for this species, it is crucial to accurately identify purebred taurines. Techniques in use today include empirical method and biological analysis. We offer in this paper a supervised Machine Learning approach of pure-bred taurine recognition. Five algorithms were trained using morphological data from hundreds of cows. Each of the models produced promising results. The RBF non linear SVM performs the best with up to 87% accuracy and 0.9308 of AUC. Furthermore, the correlation coefficients allowed to define the most discriminating morphological trait.
{"title":"An intelligent system for taurine breed recognition: preliminary results","authors":"Bembamba Fulbert, O. T. Frédéric, Malo Sadouanouan, Yougbare Bernadette, O. Dominique","doi":"10.1109/CSI54720.2022.9924139","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9924139","url":null,"abstract":"Uncontrolled crossbreeding between zebus and taurine cattle is jeopardizing the genetic heritage of West African taurines and their specific ability to resist trypanosomosis. to achieve any successful conservation policy for this species, it is crucial to accurately identify purebred taurines. Techniques in use today include empirical method and biological analysis. We offer in this paper a supervised Machine Learning approach of pure-bred taurine recognition. Five algorithms were trained using morphological data from hundreds of cows. Each of the models produced promising results. The RBF non linear SVM performs the best with up to 87% accuracy and 0.9308 of AUC. Furthermore, the correlation coefficients allowed to define the most discriminating morphological trait.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"84 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116072259","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9923956
R. Vijay, S. Nanda
In this paper, a seismicity declustering model is proposed using a sliding temporal window-based feature extraction with K-means algorithm. This approach transforms the primary features: like occurrence time, location, and magnitude of earthquake event of a catalog into the overlapping sliding window-based features (mean deviation, coefficient of variation (COV) in time and spatial domain along with the average value of magnitude). These extracted features with fewer sample sizes are used as input to the K-means algorithm for distinguishing two important classes of earthquake: aftershocks and background. This proposed method is applied to the earthquake catalog of Zagros (Iran) from the period 2006 to 2019. The simulation results revealed that three major earthquake clusters are identified in class-I which comprised of foreshock-mainshock-aftershock sequences. The events belonging to class-I have intermediate magnitude, less inter-event time (IET) & space (IED), and high COV value. The events belonging to class-II represent the characteristics of regular background seismicity (approximately 71 %) in the region. The seismicity characteristics are reported in the form of epicenter plot, space-time diagram, IET vs IED scatter plot” and other statistical values like the Silhouette index.
{"title":"Sliding Temporal Window-based Feature Extraction with K-means Clustering for Zagros (Iran) Seismicity Analysis","authors":"R. Vijay, S. Nanda","doi":"10.1109/CSI54720.2022.9923956","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9923956","url":null,"abstract":"In this paper, a seismicity declustering model is proposed using a sliding temporal window-based feature extraction with K-means algorithm. This approach transforms the primary features: like occurrence time, location, and magnitude of earthquake event of a catalog into the overlapping sliding window-based features (mean deviation, coefficient of variation (COV) in time and spatial domain along with the average value of magnitude). These extracted features with fewer sample sizes are used as input to the K-means algorithm for distinguishing two important classes of earthquake: aftershocks and background. This proposed method is applied to the earthquake catalog of Zagros (Iran) from the period 2006 to 2019. The simulation results revealed that three major earthquake clusters are identified in class-I which comprised of foreshock-mainshock-aftershock sequences. The events belonging to class-I have intermediate magnitude, less inter-event time (IET) & space (IED), and high COV value. The events belonging to class-II represent the characteristics of regular background seismicity (approximately 71 %) in the region. The seismicity characteristics are reported in the form of epicenter plot, space-time diagram, IET vs IED scatter plot” and other statistical values like the Silhouette index.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"19 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120901606","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9924050
Pallavi. M. O, S. N, M. Sundaram, Preetham N
Handwriting recognition is one of the foremost areas of exploration in pattern identification and recognition. as of now we have a system recognizes that converts the handwritten character into the printed text of many languages like English, Kannada, Tamil, Bengali, Latin, Devanagari, etc., and also the system which converts handwritten, printed copy, image or other documents into digitized format. The objective of the research is to identify the originality of handwritten documents using deep learning methods. During pandemics much of the offline work is shifted to online work, some of them are education, banking, etc. The originality of handwritten copies is checked for fraud copies to genuine verification from the original owner copies. The research includes a collection of sample written documents of around 1000 characters which consist of all the possible characters and numbers called training data, later it is cross verified by the testing data with a new input document. The framework includes proposing a CNN model and feature extraction using neural networks and proves the originality of the written copy against the test written copy. The steps involved are pre-processing, followed by segmentation, in turn, feature extraction, recognition, and comparison of each word, stroke, height, and slant of the alphabet to verify with test input. The data set will be pre-processed, CNN model extracts the features of each character, and generates a threshold value, which is compared with the testing data threshold values, if the result returned is more than 90% document will be considered to be accepted as original handwritten of the individual and if the threshold value comparison fails and less than 90% matching, the script/document is rej ected, it is categorized to fraud document.
{"title":"Deep Learning Based Application in Identifying Originality of the Hand Written Document using Convolution Neural Network","authors":"Pallavi. M. O, S. N, M. Sundaram, Preetham N","doi":"10.1109/CSI54720.2022.9924050","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9924050","url":null,"abstract":"Handwriting recognition is one of the foremost areas of exploration in pattern identification and recognition. as of now we have a system recognizes that converts the handwritten character into the printed text of many languages like English, Kannada, Tamil, Bengali, Latin, Devanagari, etc., and also the system which converts handwritten, printed copy, image or other documents into digitized format. The objective of the research is to identify the originality of handwritten documents using deep learning methods. During pandemics much of the offline work is shifted to online work, some of them are education, banking, etc. The originality of handwritten copies is checked for fraud copies to genuine verification from the original owner copies. The research includes a collection of sample written documents of around 1000 characters which consist of all the possible characters and numbers called training data, later it is cross verified by the testing data with a new input document. The framework includes proposing a CNN model and feature extraction using neural networks and proves the originality of the written copy against the test written copy. The steps involved are pre-processing, followed by segmentation, in turn, feature extraction, recognition, and comparison of each word, stroke, height, and slant of the alphabet to verify with test input. The data set will be pre-processed, CNN model extracts the features of each character, and generates a threshold value, which is compared with the testing data threshold values, if the result returned is more than 90% document will be considered to be accepted as original handwritten of the individual and if the threshold value comparison fails and less than 90% matching, the script/document is rej ected, it is categorized to fraud document.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126651185","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9924125
Pawan Kumar, Nilanjan Chattaraj
In rotational motion, there are several physical phenomena, which exclusively exist only if the angular speed varies. Those phenomena disappear at constant angular speed in rotating objects since constant angular speed establishes a static state through its constant centrifugal force. The phe-nomena such as (a) RPM-variation induced stress-variation inside rotating blades (b) RPM-variation induced low-frequency vibration generation inside rotating elements (c) RPM-variation induced energy harvesting inside rotating objects (d) RPM-variation induced variable vortex formation can fall under the mentioned category, which requires continuous and synchronized monitoring of RPM-variation in correlation with the mentioned phenomena. Firstly, commercially available typical RPM meters, which provide discrete angular speed measurements do not satisfy this requirement. Secondly, to capture the dynamical behavior, those phenomena require real-time, continuous and synchronized RPM-variation monitoring preferably through a cyber-physical connectivity for the emerging loT systems. Therefore, this paper presents the design and implementation of a cyber-physical RPM variometer featuring real-time, continuous, synchronized data-acquisition using MQTT protocol. The dashboard-GUI of the measurement system displays the RPM-tracing in the local-terminal, as well as, in the remote-terminal. The interface provides a configurable and interactive platform for real-time RPM variation measurement with the facility of measurement-parameter customization. The measurement system operates within a range of 1 to 15000 RPM with a minimum accuracy of 99.5 % for a rated scanning time of 2 sec, which is customizable. The developed non-contact type measurement system provides the facility of integrability with several IoT-enabled hardware peripherals.
{"title":"A Cyber-Physical RPM Variometer using MQTT Protocol for Real-time Continuous Data-Acquisition","authors":"Pawan Kumar, Nilanjan Chattaraj","doi":"10.1109/CSI54720.2022.9924125","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9924125","url":null,"abstract":"In rotational motion, there are several physical phenomena, which exclusively exist only if the angular speed varies. Those phenomena disappear at constant angular speed in rotating objects since constant angular speed establishes a static state through its constant centrifugal force. The phe-nomena such as (a) RPM-variation induced stress-variation inside rotating blades (b) RPM-variation induced low-frequency vibration generation inside rotating elements (c) RPM-variation induced energy harvesting inside rotating objects (d) RPM-variation induced variable vortex formation can fall under the mentioned category, which requires continuous and synchronized monitoring of RPM-variation in correlation with the mentioned phenomena. Firstly, commercially available typical RPM meters, which provide discrete angular speed measurements do not satisfy this requirement. Secondly, to capture the dynamical behavior, those phenomena require real-time, continuous and synchronized RPM-variation monitoring preferably through a cyber-physical connectivity for the emerging loT systems. Therefore, this paper presents the design and implementation of a cyber-physical RPM variometer featuring real-time, continuous, synchronized data-acquisition using MQTT protocol. The dashboard-GUI of the measurement system displays the RPM-tracing in the local-terminal, as well as, in the remote-terminal. The interface provides a configurable and interactive platform for real-time RPM variation measurement with the facility of measurement-parameter customization. The measurement system operates within a range of 1 to 15000 RPM with a minimum accuracy of 99.5 % for a rated scanning time of 2 sec, which is customizable. The developed non-contact type measurement system provides the facility of integrability with several IoT-enabled hardware peripherals.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"119 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126866073","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9923976
Yash Jha, Harsh Prajapati, B. Fataniya
Object detection has been evolving greatly in recent years and the advancements in hardware and software technologies have made it possible to perform object detection with ease. Due to the enhanced capabilities of the modern processors and Graphics Processing Unit (GPU) of doing an exponentially complex and extensive number of iterations in very less time. Real-time object detection has become highly popular and the center of attention in recent years because most of the hardware owned by common users is powerful enough to compute that which unlocks whole new possibilities for implementing real-time object detection in numerous applications in various domains. Real-time herbal plant detection is one such topic that has many applications in the field of ayurvedic medicines and many other pharmaceutical applications that can be used to spike the efficiency in identifying these herbal plants that can be used as a precaution and even as a cure for numerous health problems. There are many existing algorithms for real-time detection, but the evolution of new Artificial Neural Network (ANN) and Machine Learning (ML) techniques unlocks new ways to implement recent and advanced algorithms to apply for real-time detection of such powdered microscopic images to achieve better performance in various aspects compared to already existing methods. Our model is trained for detecting three types of microscopic herbal plants.
{"title":"Real-Time Object Detection in Microscopic Image of Indian Herbal Plants using YOLOv5 on Jetson Nano","authors":"Yash Jha, Harsh Prajapati, B. Fataniya","doi":"10.1109/CSI54720.2022.9923976","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9923976","url":null,"abstract":"Object detection has been evolving greatly in recent years and the advancements in hardware and software technologies have made it possible to perform object detection with ease. Due to the enhanced capabilities of the modern processors and Graphics Processing Unit (GPU) of doing an exponentially complex and extensive number of iterations in very less time. Real-time object detection has become highly popular and the center of attention in recent years because most of the hardware owned by common users is powerful enough to compute that which unlocks whole new possibilities for implementing real-time object detection in numerous applications in various domains. Real-time herbal plant detection is one such topic that has many applications in the field of ayurvedic medicines and many other pharmaceutical applications that can be used to spike the efficiency in identifying these herbal plants that can be used as a precaution and even as a cure for numerous health problems. There are many existing algorithms for real-time detection, but the evolution of new Artificial Neural Network (ANN) and Machine Learning (ML) techniques unlocks new ways to implement recent and advanced algorithms to apply for real-time detection of such powdered microscopic images to achieve better performance in various aspects compared to already existing methods. Our model is trained for detecting three types of microscopic herbal plants.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114308460","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9923999
E. Kadusic, C. Ruland, Narcisa Hadzajlic, N. Zivic
The designing process of an IoT (Internet of Things) network requires adequate knowledge of various communication technologies that make the connection of the IoT modules possible. Many important factors such as scalability, bandwidth, data rate (speed), coverage, power consumption, and security support need to be considered to answer the needs of an IoT application with regards to the implemented radio communication technology. This paper studies the choices of three major LPWAN (Low-Power Wide-Area Networks) technologies that are currently leading in the market of radio communication technologies. Focusing on Sigfox, LoRaWAN (Low-Range Wide-Area Networks), and NB-IoT (Narrow-Band Internet of Things), this work intends to give the respective pros and cons of the mentioned technologies and a clear view of the recent trends and effective choices of radio communication technologies for major smart IoT applications.
{"title":"The factors for choosing among NB-IoT, LoRaWAN, and Sigfox radio communication technologies for IoT networking","authors":"E. Kadusic, C. Ruland, Narcisa Hadzajlic, N. Zivic","doi":"10.1109/CSI54720.2022.9923999","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9923999","url":null,"abstract":"The designing process of an IoT (Internet of Things) network requires adequate knowledge of various communication technologies that make the connection of the IoT modules possible. Many important factors such as scalability, bandwidth, data rate (speed), coverage, power consumption, and security support need to be considered to answer the needs of an IoT application with regards to the implemented radio communication technology. This paper studies the choices of three major LPWAN (Low-Power Wide-Area Networks) technologies that are currently leading in the market of radio communication technologies. Focusing on Sigfox, LoRaWAN (Low-Range Wide-Area Networks), and NB-IoT (Narrow-Band Internet of Things), this work intends to give the respective pros and cons of the mentioned technologies and a clear view of the recent trends and effective choices of radio communication technologies for major smart IoT applications.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123455191","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9924134
Disha S, Deekshitha B, Anwitha A, Kavyashree U M, Shrikanth Rao S.K, R. J. Martis
Atrial Fibrillation (AF) is a life-threatening heart rhythm disorder. AF diagnosis is very essential and important aspect for healthcare professionals. Early detection of AF using Electrocardiogram (ECG) plays an important role in the clinical practice. Manual interpretation of ECG signals to detect AF is time-consuming and needs higher expertise, and it is subject to variability among experts. Detecting AF in a timely and effective manner still remains a difficult challenge. In this paper, we propose a Deep Learning (DL) based AF detection method using Physionet challenge 2017 dataset. VGG16 architecture is used for the classification purpose. With the help of Discrete Wavelet Transform (DWT) the ECG signals are denoised. The RR intervals are computed and are subjected to VGG16 for classification. The class specific accuracies of normal, AF, and other rhythms are calculated. The proposed method achieves overall accuracy of 97.60%. The proposed method can be used as an assisted tool by the physician in their clinical practice.
{"title":"Deep Learning assisted tool for Atrial Fibrillation detection using RR Intervals","authors":"Disha S, Deekshitha B, Anwitha A, Kavyashree U M, Shrikanth Rao S.K, R. J. Martis","doi":"10.1109/CSI54720.2022.9924134","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9924134","url":null,"abstract":"Atrial Fibrillation (AF) is a life-threatening heart rhythm disorder. AF diagnosis is very essential and important aspect for healthcare professionals. Early detection of AF using Electrocardiogram (ECG) plays an important role in the clinical practice. Manual interpretation of ECG signals to detect AF is time-consuming and needs higher expertise, and it is subject to variability among experts. Detecting AF in a timely and effective manner still remains a difficult challenge. In this paper, we propose a Deep Learning (DL) based AF detection method using Physionet challenge 2017 dataset. VGG16 architecture is used for the classification purpose. With the help of Discrete Wavelet Transform (DWT) the ECG signals are denoised. The RR intervals are computed and are subjected to VGG16 for classification. The class specific accuracies of normal, AF, and other rhythms are calculated. The proposed method achieves overall accuracy of 97.60%. The proposed method can be used as an assisted tool by the physician in their clinical practice.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"737 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131689664","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9923952
Sakthi Vel S, P. R
Cross Language Information Retrieval (CLIR), is the process of retrieving relevant documents, where in the language of the given query is different from the language of the retrieved documents. CLIR systems allow the users to search and access documents in the language different from the language of the search query. CLIR systems have been divided into Monolingual CLIR, Bi-lingual CLIR, and Multilingual CLIR based on different languages of query and documents. The first step of the Cross Language Information Retrieval system is the text pre-processing of given text documents in to useful representations. Pre-processing is the set of tasks that convert the given text documents into a suitable format for any higher-level text related applications. This technique can be used to reduce the computational process, noise data, and irrelevant information from the given text documents. This paper discusses in detail the different pre-processing techniques such as dataset creation, tokenization, noise removal, stop word removal, stemming, lemmatization and finally term weighting of two languages dataset (i.e., Tamil and Malayalam), which is manually collected from BBC online website. Finally, the study investigates feature extraction techniques of Term Frequency- Inverse Document Frequency (TF-IDF). These techniques will help to design and model CLIR systems with high performance.
{"title":"Text Pre-Processing Methods on Cross Language Information Retrieval","authors":"Sakthi Vel S, P. R","doi":"10.1109/CSI54720.2022.9923952","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9923952","url":null,"abstract":"Cross Language Information Retrieval (CLIR), is the process of retrieving relevant documents, where in the language of the given query is different from the language of the retrieved documents. CLIR systems allow the users to search and access documents in the language different from the language of the search query. CLIR systems have been divided into Monolingual CLIR, Bi-lingual CLIR, and Multilingual CLIR based on different languages of query and documents. The first step of the Cross Language Information Retrieval system is the text pre-processing of given text documents in to useful representations. Pre-processing is the set of tasks that convert the given text documents into a suitable format for any higher-level text related applications. This technique can be used to reduce the computational process, noise data, and irrelevant information from the given text documents. This paper discusses in detail the different pre-processing techniques such as dataset creation, tokenization, noise removal, stop word removal, stemming, lemmatization and finally term weighting of two languages dataset (i.e., Tamil and Malayalam), which is manually collected from BBC online website. Finally, the study investigates feature extraction techniques of Term Frequency- Inverse Document Frequency (TF-IDF). These techniques will help to design and model CLIR systems with high performance.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128662533","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 : 2022-08-31DOI: 10.1109/CSI54720.2022.9924114
Urvi Sharma, G. Sajeev, S. S. Rani
E-Commerce has seen a lot of growth over the past decade. With an increase in commodities, especially fashion accessories and clothing items in the online-market, a need for an efficient recommendation system arises for better information filtering. Several different apparel recommendation systems already exist in the literature. However, as time passes, new challenges are arising, such as computational complexity and an exponential increase in data. Also, due to fast-changing trends, the recommendation model is required to update frequently. This work proposes an improvised collaborative-filtering based recommendation system. A ranking algorithm, Nearest Neighbor PageRank (NNPR), is developed that uses the nearest neighbors of the user along with the PageRank algorithm to generate personalized recommendations. The proposed model, is evaluated in comparison with Alternating Least Square (ALS) algorithm. The experiments are conducted on Amazon Fashion Review Dataset, and the results of this experiment are recorded in Hit-Rate (HR) and Mean-Reciprocal Ranking (MRR). It is observed, that NNPR performs better than ALS in both Active User and Cold Start scenarios. Moreover, the hybrid model ALSNNPR improves the performance of ALS using NNPR as a ranking algorithm.
{"title":"Personalized Fashion Recommendation Using Nearest Neighbor PageRank Algorithm","authors":"Urvi Sharma, G. Sajeev, S. S. Rani","doi":"10.1109/CSI54720.2022.9924114","DOIUrl":"https://doi.org/10.1109/CSI54720.2022.9924114","url":null,"abstract":"E-Commerce has seen a lot of growth over the past decade. With an increase in commodities, especially fashion accessories and clothing items in the online-market, a need for an efficient recommendation system arises for better information filtering. Several different apparel recommendation systems already exist in the literature. However, as time passes, new challenges are arising, such as computational complexity and an exponential increase in data. Also, due to fast-changing trends, the recommendation model is required to update frequently. This work proposes an improvised collaborative-filtering based recommendation system. A ranking algorithm, Nearest Neighbor PageRank (NNPR), is developed that uses the nearest neighbors of the user along with the PageRank algorithm to generate personalized recommendations. The proposed model, is evaluated in comparison with Alternating Least Square (ALS) algorithm. The experiments are conducted on Amazon Fashion Review Dataset, and the results of this experiment are recorded in Hit-Rate (HR) and Mean-Reciprocal Ranking (MRR). It is observed, that NNPR performs better than ALS in both Active User and Cold Start scenarios. Moreover, the hybrid model ALSNNPR improves the performance of ALS using NNPR as a ranking algorithm.","PeriodicalId":221137,"journal":{"name":"2022 International Conference on Connected Systems & Intelligence (CSI)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121720955","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：