Clustering of JSON document using Derived X-Path Method
In this modern era where data creation is in abundance, handling that data is the biggest question researchers have been trying to solve for the past couple of decades. The data in any form needs to be stored in such a way that the retrieval and manipulation of the data are easier and simpler. In this paper, we look at how huge datasets in JSON format are being analyzed and made into clusters for an easier understanding of the documents that we are handling. This process is done by the first pattern analyzing the documents followed by clustering those documents one by one based on similarity of the attributes (arrays and sub-documents included). The clusters are then evaluated on an algorithm designed specifically for this purpose. To conclude, this paper lays out the model on how to cluster huge JSON documents with varying levels and types.
Published by: Kaushik I.
Author: Kaushik I.
Paper ID: V7I4-1473
Paper Status: published
Published: July 22, 2021
Importance of Feature Selection in Model Accuracy
As a dimensionality reduction strategy, feature selection attempts to select small set of most important features from primary features by eliminating obsolete, data-redundant and non-relevant noisy features. This process of choosing a set of the original variables such that a model based on data containing only these features has the simplest output is known as feature selection. Feature Selection eliminates over-fitting, increases model efficiency by removing redundant functions, and has the added benefit of maintaining the primary feature representation, resulting in improved accuracy. Good learning efficiency, results into higher machine learning model accuracy, lower cost of computation, and efficient model accuracy, is typically the product of feature selection. Recently, researchers in the area of computer vision, deep Learning, data mining, and other fields have shown that several feature selection algorithms resulted in the efficiency in their work through computational theory and research. This paper aims to examine the importance of feature selection in model accuracy. Feature selection is critical for various reasons, which include simplicity, performance, computational efficiency, and accuracy. It is often used in both supervised and unsupervised learning scenarios. These strategies can help boosting the productivity of various machine learning algorithms, as well as coaching. Feature selection decreases learning time and increases data consistency and comprehension.
Published by: Pranjal Rawat, Nitin, Sameer Dev Sharma
Author: Pranjal Rawat
Paper ID: V7I4-1470
Paper Status: published
Published: July 22, 2021
College Kendra
The android app “College Kendra” is a package of unique features that a university/college need in their college life. There are 8 features in this android app namely Online Attendance, E-Passbook, Notice Board, Calculator, Anonymous Feedback, Automatic Silent Mode and Distress Alert. These might seem pretty common but they aren’t. This paper describes why in the coming sections. The main feature in this app, the Online attendance is made to save professor’s time on trivial attendance task. The amount of average time saved on daily basis will be 5 minutes. This is the time which will be wasted on taking attendance manually, which accounts over 10 hours per semester and 20 hours per year. Also, through this app we can 100% ensure that the student is really present in the class, thus eliminating proxies. What makes it completely efficient is the One-Time-Password based authentication (OTP). This particular Attendance feature totally eliminates the paper work too. With effective usage, any institute can apply this feature for conducting quick attendance and get better results in less time. The calculator feature I developed isn’t any typical or regular calculator that everyone uses. It doesn’t have the same usual operations. As graduation students use this, it’ll have prime and big-league operations such as bit convertor. The other features help the students with college tasks and non-academic problems. The E-Passbook, Notice Board and Chat Room features included in the app will help with classroom tasks and work as communication media. The Anonymous feedback feature helps to report any issue in the classroom. The Automatic Silent Mode Turner will set the phone in the silent mode automatically when you are in the class. The Distress Alert option would help students to alert others during emergency or unusual situations with just a single click. This paper gives a complete understanding about all the features of our app, which is a complete means of help for a university/college.
Published by: Pranav Kongara
Author: Pranav Kongara
Paper ID: V7I4-1469
Paper Status: published
Published: July 22, 2021
Detection of macro based attacks in office documents using Machine Learning
With the rapid developments in internet, users share information through document files generated through online or offline office software. Due to implicit trust on the web and wide acceptance of these document files (such as PDF, DOC, Office Open XML), users share documents on the web by trusting third-party services which can be easily exploited by cybercriminals to inject malicious code (Malware) into the document files by various means of exploitations. These exploitations are undetectable and easily evaded on antivirus software which makes the problem of malware detection and classification even more complex. In recent years, the attacks that leverage office documents have gradually increased and thus harder to detect since malware authors use various ways to inject malicious code on to the office documents. They offer flexibility in document structure with numerous features for attackers to exploit. In this paper, a broad classification of macro based malicious document attack is provided along with a detailed description of the attack opportunities available using office documents. A hybrid malware analysis technique is proposed which thoroughly analyze the file for any macro attacks along with decision paradigms such as machine learning is used to detect and classify the malicious document present in Microsoft Office applications such as Word, Excel, Power point.
Published by: Aishwarya, Bhumica B., Sumit Suman, Ravi V.
Author: Aishwarya
Paper ID: V7I4-1460
Paper Status: published
Published: July 22, 2021
Comparison of the presence of Comorbidities in Recovered and Expired patients with COVID-19 infection
Background The Covid 19 pandemic, has been reverberating across the globe with varying severity & disease presentations. The epidemiological factors associated with the disease included the age & the associated comorbidities like Hypertension, Diabetes Mellitus & Chronic heart & Lung diseases amongst others. Objectives To study the presence of Comorbidities in Covid 19 patients. Methods A retrospective observational study from the laboratory and medical records was conducted on consecutive 100 recovered & Expired patients admitted to a tertiary care centre. Results & Conclusions; The average age of the Recovered & Expired patients was 43.68 years & 58.71 years with a male female ratio of 58:42 & 74::26 respectively. An increase in the age with an increasing no of comorbidities was seen in both the groups. A significant male preponderance was seen showing an increase with increasing number of the associated comorbidities. Amongst other comorbidities, more than 50% of the expired patients were hypertensive compared to only 20% of the recovered patients. 38% of the expired patients were admitted with a history of Diabetes Mellitus compared to 23% of the expired patients. 70% of all patients having Hypertension & 63% of all having DM as the only comorbidity expired from the disease. 86% of all patients with a cardiac disease & 100% of all patients in the study having an associated kidney disease or immunocompromised expired from the disease. Hence a significant association was found between associated comorbidities & mortality in the patients.
Published by: Dr. Anjali Goyal, Dr. Heta Patel, Dr. Chandni Gadara, Dr. Shachi Shah, Dr. Misha Antani
Author: Dr. Anjali Goyal
Paper ID: V7I4-1425
Paper Status: published
Published: July 22, 2021
Fault Detection and Diagnosis in PV System Array
Solar energy has procurement increasing attention in modernistic year. The power generation from solar source is always clean and free from pollution. In recent year, the price of solar PV panel is going downwards which increased the attention to use solar PV application in the recent years. Machine Learning techniques are employed in photovoltaic (PV) systems to make full utilization of PV array output power which depends on solar irradiation and ambient temperature. For such a large scale growing business it is necessary for it to be free of faults so that are ensured smooth and hassle free working.
Published by: Vinit Koli, Abhijit Shankar Iyer, Aniket Kedare, Vinayak More, Milind Paraye
Author: Vinit Koli
Paper ID: V7I4-1284
Paper Status: published
Published: July 22, 2021
