• Send Your paper at :
  • ijirtm@gmail.com
  • editor@ijirtm.com

Impact Factor - 5.445
Impact Factor - 5.68



IJIRTM: Volume-5, Issue-5, 2021

Paper Title : A Method To Improve Fake News Detection Using Machine Learning Algorithm on Social Media
Author Name : Mojahidul Islam, Chinmay Bhatt, Varsha Namdeo
Keywords : fake News Detection, Scar Scam, Rumor, AI, Machine Learning.
Abstract :

News is a vital piece of our life. In everyday life, current news is useful to upgrade information that occurs throughout the planet. So the majority of people groups incline toward watching news the greater part of the people groups by and large favor perusing paper promptly toward the beginning of the day appreciating with a cup of tea. If the news is fake that will misdirect people groups in some cases fake words used to get out reports about things or it will influence some political pioneer positions as a result of fake news. This is the reason we proposed a fake news detection framework, however presently the measure of everyday information on the Internet or interpersonal organizations is developing quickly, recognizing fake news or not looking through every one of the information is insane to the point that it consumes a large chunk of the day to utilize a colossal arranging technique to This work proposes a grouping based fake news recognition framework, like NaiveBayes (NB), Support Vector Machine (SVM), K Nearest Neighbor (KNN), and Decision Tree (DT). We thought about all the AI strategies used to identify fake news. Simulation is performed using Python Spyder 3.6 software. Results show that the proposed decision tree method achieves the maximum accuracy that is 98% and the error rate is 2% while the existing approach achieved 91% accuracy and 9% error rate. Thus the comparison of the previous and proposed method in terms of methodology, accuracy, and error rate is done. Therefore is clear from optimized results that the proposed method is giving better performance than the previous method.

Download IJIRTM-0505202101
Paper Title : A Survey on Structural Planning, Scheduling and Resource Allocation of Project
Author Name : Akash Mishra, Kapil Malviya
Keywords : Project Management, Resource Scheduling, Resource Leveling, Structural Planning.
Abstract :

A few development exercises can be figured out how to accomplish the benefit inside restricted assets and time. In this manner project the board procedures are helpful in planning and organizing the different assets by controlled technique. The board strategies like Critical Path Method, Program Evaluation and Review Techniques (CPM/PERT) have been effectively executed before the 1970's, in different Civil Engineering projects in the nations like USA, Canada, Australia. These procedures help the executives in proficient and financial utilization of assets for consummation of venture goals with limitless accessibility of assets, however it is seen that assets are restricted in genuine task situation. It has been seen that the undertaking delays happen because of inadequate stockpile of assets. In huge scope projects, setting up an exact and useful arrangement is truly challenging. PC bundles like MS Project and Primavera project organizer are utilized in development industry. Task the board methods can be utilized to determine asset clashes and furthermore helpful in limiting the undertaking term inside restricted accessibility of assets to make the venture beneficial. The primary point of this review is to break down the Project the executives’ methods by booking different development exercises, distribution of assets and asset evening out utilizing Microsoft Project 2013 for private structure. This paper dissects asset compelled project utilizing Microsoft Project 2013 by asset evening out and contrasts the time cost suggestions and booked time and assessed cost.

Download IJIRTM-0505202102
Paper Title : A Review on Fake News Detection Methods Using Machine Learning
Author Name : Mojahidul Islam, Chinmay Bhatt, Varsha Namdeo
Keywords : Fake News Detection, Machine Learning, Deep Learning, Rumor, Hoax.
Abstract :

Social media is one of the most accessible news sources these days for some people worldwide because of their low worth, speedy access, and quick spread. Nonetheless, this accompanies some befuddling signs and huge dangers of openness to 'bogus stories' composed to misdirect per users. Such data can influence the public's voice and permit fiendish gatherings to control the result of public occasions, like races. Phony and deluding news can truly affect the individuals who wind up as targets. This paper centers on investigation of 2017 to 2021 papers and examination of various phony news discovery strategies. This overview gives a broad audit about the new and past assessments on bogus news identification utilizing diverse Machine Learning calculations.

Download IJIRTM-0505202103
Paper Title : Twitter Data Extraction Method for Community Detection on Social Media
Author Name : Amit Kumar Pandey, Chinmay Bhatt, Varsha Namdeo
Keywords : Twitter Dataset, Data Extraction, Social Media, Community Detection, Social Networks.
Abstract :

In the present situation web-based media is an arising field for some analysts. In web-based media the information created through client side is colossal. To keep up with the client produced information there are many mining errands are available in online media mining. There are numerous long range interpersonal communication destinations where client makes their own local area based on their advantage. As it is realized that web-based media is a major virtual world in that numerous clients have their profile and they are associated with various sort of gatherings. Community discovery is the main element of online media. It is like the grouping element of information mining. In part based local area recognition a ton of work has been done, yet distinguishing local area through impact is an alternate method of identification in online media mining. The primary goal of the exposition is to recognize the local area taking advantage of Leverage. Major issues of community detection are scalability and quality of the community. Some of the algorithm scalable in large network and provides better results as compare to other algorithm. We have compared the algorithms on the social data of the twitter. As result it is prove that algorithms are scalable in the large network as per the evaluation parameter. The unique feature of this thesis is that we have evaluated all the features of the algorithm on the large social network.

Download IJIRTM-0505202104
Paper Title : A Novel Analysis Method For The Effect of Coefficient of Performance (Cop) on Solar Refrigerator With Axial D.C. Fans
Author Name : Kapil Khandelwal, Rupanshu Suhane, M. K. Chopra
Keywords : D C Fan, Solar Panel, Solar Energy, COP.
Abstract :

A domestic fridge is utilized practically in each home for safeguarding food things, cooling the water, and some more. A cooler has primarily four parts that run the fridge and these are the blower, condenser, choking valve, and evaporator. This test examination is done on a homegrown fridge controlled by a sunlight-powered charger. This examination is explored on a cooler with additional fans fitted on the backboard of the fridge. These fans were the primary alteration in this analysis since it works on the presentation of the cooler. Crafted by the fans is to cool the condenser (cylinder and wire network). This cooling is finished by constrained air cooling which isn't finished by the normal air cooling in a straightforward fridge. The air conveys the warmth of the condenser and moved it to the climate. The fans utilized here are hub fans. These fans were worked by an immediate current which is provided by the battery. In this test, the refrigerant which is coursing through the cooler chills off by fans thus the temperature lets down. This refrigerant is extended (by the choking gadget) and afterward, the temperature of the refrigerant diminished to a lot lesser worth in contrast with the cooler which is worked without fans. Then, at that point, this lower temperature refrigerant went to the evaporator. The evaporator removes heat from the food things and cools the food things in lesser time because of this additional decrement in temperature. In this exploratory arrangement refrigerant, R134a is utilized in the fridge. Here the solar-powered charger is utilized to run the fridge. The solar-powered charger is a sustainable wellspring of energy. An inexhaustible wellspring of energy is that energy that can be recharged over and over. It never closes because solar-based energy is accessible in an exceptionally large sum. The energy utilized in this examination is without contamination. This work will build the reliance on sustainable power since it just takes a one-time venture and the support is less. In this test set-up, execution tests were done under the controlled surrounding condition. At the point when the fridge is worked without a fan the cooling got is less and the cooling got by the fan-worked cooler is extremely high. The fridge without fans sets aside more effort to changes water into ice in examination over to the fans-worked cooler in the cooler. So this fan-worked cooler gives excellent outcomes. There is a lot of contrast between fridges worked with fans and without fans. The outcome shows that the homegrown fridge with D.C. pivotal fans gives a superior coefficient of execution (COP) in contrast with the basic homegrown cooler. This outcome shows that the normal coefficient of execution (COP) of the homegrown fridge with D.C. pivotal fans was 0.038 higher than the basic homegrown cooler in the concentrated range. It is seen that for the main hour COP was lower for the cooler with a fan yet it then, at that point expanded for the remainder of the time. The exploratory examination result shows the refrigeration impact for the cooler with the fan was about 7.02% higher than the fridge without a fan. In this exploratory perception, the cooler lodge temperature was additionally diminished for the fridge with a fan. Additionally, the condenser temperature was lesser for the cooler with a fan and higher without a fan. So this test gives the further developed presentation of the fridge. After the change, it upgrades the coefficient of execution just as the refrigeration impact.

Download IJIRTM-0505202105
Paper Title : Design and Strength Check For Turbine Inlet Butterfly Valve Using Computational Testing Method
Author Name : Akhil Kumar, Prof.Gagan Varshney
Keywords : Butterfly Valve, FEM, CAD design, Stress, Deformation, CATIA, ANSYS.
Abstract :

Butterfly valves are used to control discharge of fluids in penstock of hydropower plants or industrial pipe networks. It has a disc installed in between which can be made to rotate manually or automatically by pneumatic servomotors. These valves are also used as complete shut off valve. The disc of butterfly valve is subjected to pressure of fluid flowing in pipes which tries to deform it. The resistance to this pressure is offered by disc when stresses are induced in it. For safe working of butterfly valves, it is necessary that the stresses induced do not exceed elastic limit otherwise it will lead to its permanent deformation. The 3D modeling to be performs for butterfly valve by using CAD software. Further the stress & displacement FEM analysis of the butterfly valve to be performed by using ANSYS tool to evaluate the optimized result.

Download IJIRTM-0505202106
Paper Title : A Method for Rumor Detection on Social Media using Bi-LSTM Algorithm
Author Name : Swati Goswami, Prof.Chetan Agrawal, Bhavana Verma
Keywords : Deep Learning, Rumor Detection, RNN, BI-LSTM, Social Media.
Abstract :

The momentum of the Internet, the fast advancement of cell phones, the accessibility to all, and the advantages of simple correspondence and association, prompted facilitating dispersal of news and data people via web-based networking media systems without control of the substance gave through these destinations. This represents a genuine hazard to the truthiness of web based distributing. The misleading data dispersed through online life destinations spreads quickly and causes negative impacts in different regions. In this work, we have implemented a Bi-Directional function with LSTM for optimizing the model. The bi-Directional function is the way by which we can pass both direction tasks as a function. The texts are preprocessed based on function later on the preprocessed data gets classified using LSTM. The proposed work will be compared based on 4 parameters that are accuracy, precision, recall, and F1 score. The accuracy of the proposed method is almost 85% and other parameters are also enhanced in the proposed method.

Download IJIRTM-0505202107
Paper Title : Spectrum Sensing in Cognitive Radio Network: Survey and Discussions
Author Name : Neha Kapse, Prof.Jitendra Mishra
Keywords : Cognitive radio, Deep learning, spectrum sensing, Convolutional neural network.
Abstract :

The advent of new applications and technologies such as the Internet of Things, Cyber-Physical Systems, etc., has propelled the demand for wireless spectrum [1]. This increase in demand for spectrum cannot be achieved easily as a spectrum is a limited resource, and its expansion is difficult due to technological limitations. Cognitive radio is a promising technology which allows secondary users (SUs) to access the licensed band of primary user (PU) opportunistically when it is not being used by the primary user. Thus, the transmission of the PU is not impacted in any way. In this paper we review the different techniques in cognitive radio for the spectrum sharing, in future try to enhance the utilization of unused channel and frequency for secondary user.

Download IJIRTM-0505202108
Paper Title : Inhabited Community Detection Using Improved K-Means Clustering Algorithm
Author Name : Kanishka Sisodia, Chinmay Bhatt, Varsha Namdeo
Keywords : K Means, Clustering, Community Detection, Machine Learning.
Abstract :

In clustering objects those have comparative nature will lies in a similar cluster and on the off chance that they are of unmistakable nature, they will be in various cluster. However Standard K-means is prime calculations of the clustering yet it experience the ill effects of certain detriments these are as per the following 1) Performance relies upon introductory groups which are picked haphazardly in standard K-means. 2) The standard K-means calculation has time intricacy of 0(nkl) that is an excessive amount of costly. 3) The standard K-means calculation likewise experiences the dead unit issue those outcomes in clusters without any information focuses. 4) In standard K-means we do arbitrary instatement which makes them chat at nearby minima. Numerous upgrades were proposed to work on the presentation of the standard K-means calculation yet the vast majority of them address just each of them in turn. In this paper, we address introductory focus just as calculation intricacy issue in one calculation. Now a day, population growth rate increase rapidly. So the size of the population database is increased exponentially. It is very difficult to find information from this huge dataset. Both clustering and classification algorithm is used to extract data from population database. The size of family, Population Density, Birth Rate, Death Rate, number of Employed person, Unemployment, Prediction of Male person, Prediction of Female population, Prediction of Budget for the year, Prediction of members in each caste, Prediction of Rural population and Prediction of Urban population etc.

Download IJIRTM-0505202109
Paper Title : A Methodology for Secured Energy Efficient Routing Protocol for Wireless Sensor Network
Author Name : Tanvi Tomar, Chinmay Bhatt, Varsha Namdeo
Keywords : WSN, Energy Efficient Protocols, Routing, Security, QoS.
Abstract :

The dramatically request of Quality of Service (QoS) correspondence has made remote correspondence as an unavoidable advances. Among major within reach advances, Wireless Sensor Network (WSN) has been tracked down a possible answer for meet significant genuine prerequisites, including checking and control, observation, medical services, traffic reconnaissance and protection frameworks. Working with QoS requests, energy-productivity and secure correspondence has consistently been an open examination region for the scholarly community ventures. The absence of safety powers organization to go through compromised circumstance as well as makes it energy thorough. Then again, giving deferral versatile, energy-productive, higher all through and secure correspondence makes WSN strong and proficient. Notwithstanding, empowering vigorous arrangement with ideal security methods related to improved directing model is an open exploration region. With this inspiration, in this proposal a portion of the methods and directing methodologies for energy proficient and secure correspondence with improved steering convention can guarantee secure and energy-productive correspondence over WSNs. an energy-productive directing convention with information transmission security for wireless Sensor organizations. We make an energy and distance mindful sink-established tree in the organization which is utilized for secure information transmissions from the source sensors to the base station and pre-stacked shared mystery keys.

Download IJIRTM-0505202110