Showing papers by "National Institute of Technology, Karnataka published in 2020"
16 Dec 2020
TL;DR: The HASOC track as mentioned in this paper is dedicated to evaluate technology for finding offensive language and hate speech, which has attracted much interest and over 40 research groups have participated as well as described their approaches in papers.
Abstract: This paper presents the HASOC track and its two parts. HASOC is dedicated to evaluate technology for finding Offensive Language and Hate Speech. HASOC is creating test collections for languages with few resources and English for comparison. The first track within HASOC has continued work from 2019 and provided a testbed of Twitter posts for Hindi, German and English. The second track within HASOC has created test resources for Tamil and Malayalam in native and Latin script. Posts were extracted mainly from Youtube and Twitter. Both tracks have attracted much interest and over 40 research groups have participated as well as described their approaches in papers. In this overview, we present the tasks, the data and the main results.
127 citations
TL;DR: In this article, a review of 2D transition metal carbides, carbonitrides and nitrides in conjunction with surface terminations namely fluorine, hydroxyl or oxygen which add hydrophilicity to their surfaces is presented.
116 citations
01 Jan 2020
TL;DR: In this paper, the authors adopted the Analytical Hierarchy Process (AHP) and Multi influence factor (MIF), multi-criteria decision-making approaches to determine the weights for the influencing factors.
Abstract: Groundwater is one of the most vital natural resources; spatially varying in quality and quantity. Increased urbanisation and population creates tremendous pressure on the quality and quantity of the groundwater resources. In this study, Ponnaniyaru watershed of Cauvery basin was considered for this research. Geographical information system (GIS) and remote sensing (RS) plays a vital role in preparing various thematic layers for targeting the groundwater potential zones (GWPZ). This study adopts the Analytical Hierarchy Process (AHP) and Multi influence factor (MIF), multi-criteria decision-making approaches to determine the weights for the influencing factors. Weighted linear overlay analysis was carried out to determine the GWPZ. Further, the resultant GWPZ map has been reclassified into five different classes, namely Very good, Good, Moderate, Poor and Very poor. The results were validated with observed well-yield data, and the predictive precision for AHP and MIF was found to be 75%, and 71% respectively.
82 citations
TL;DR: The presented IPUC inverter has low voltage stress on switches and is capable of voltage boosting, and a new voltage balancing method based on logic form equations is developed for regulating the inherent floating capacitor voltage to half the input dc voltage.
Abstract: In this brief, a seven-level (7L) improved packed U-cell (IPUC) inverter with reduced power electronic components is proposed. The presented IPUC inverter has low voltage stress on switches and is capable of voltage boosting. A new voltage balancing method based on logic form equations is developed for regulating the inherent floating capacitor voltage to half the input dc voltage. The proposed 7L IPUC is compared with other state-of-the-art 7L inverters in terms of number of IGBTs, blocking voltage, and driver circuits for attesting its superior merits. The performance of the proposed voltage balancing is verified through a laboratory prototyped 7L IPUC inverter considering varying load conditions and the corresponding results are elucidated.
80 citations
TL;DR: This paper elucidates on the way of extracting email content and behavior-based features, what features are appropriate in the detection of UBEs, and the selection of the most discriminating feature set, and facilitates an exhaustive comparative study using several state-of-the-art machine learning algorithms.
Abstract: With the influx of technological advancements and the increased simplicity in communication, especially through emails, the upsurge in the volume of unsolicited bulk emails (UBEs) has become a severe threat to global security and economy. Spam emails not only waste users’ time, but also consume a lot of network bandwidth, and may also include malware as executable files. Alternatively, phishing emails falsely claim users’ personal information to facilitate identity theft and are comparatively more dangerous. Thus, there is an intrinsic need for the development of more robust and dependable UBE filters that facilitate automatic detection of such emails. There are several countermeasures to spam and phishing, including blacklisting and content-based filtering. However, in addition to content-based features, behavior-based features are well-suited in the detection of UBEs. Machine learning models are being extensively used by leading internet service providers like Yahoo, Gmail, and Outlook, to filter and classify UBEs successfully. There are far too many options to consider, owing to the need to facilitate UBE detection and the recent advances in this domain. In this paper, we aim at elucidating on the way of extracting email content and behavior-based features, what features are appropriate in the detection of UBEs, and the selection of the most discriminating feature set. Furthermore, to accurately handle the menace of UBEs, we facilitate an exhaustive comparative study using several state-of-the-art machine learning algorithms. Our proposed models resulted in an overall accuracy of 99% in the classification of UBEs. The text is accompanied by snippets of Python code, to enable the reader to implement the approaches elucidated in this paper.
79 citations
TL;DR: Experimental results demonstrate that the proposed HYBRID algorithm outperforms peer research and benchmark algorithms in terms of efficient utilization of the cloud resources, improved reliability and reduced average response time.
Abstract: In this paper, we propose a novel HYBRID Bio-Inspired algorithm for task scheduling and resource management, since it plays an important role in the cloud computing environment. Conventional scheduling algorithms such as Round Robin, First Come First Serve, Ant Colony Optimization etc. have been widely used in many cloud computing systems. Cloud receives clients tasks in a rapid rate and allocation of resources to these tasks should be handled in an intelligent manner. In this proposed work, we allocate the tasks to the virtual machines in an efficient manner using Modified Particle Swarm Optimization algorithm and then allocation / management of resources (CPU and Memory), as demanded by the tasks, is handled by proposed HYBRID Bio-Inspired algorithm (Modified PSO + Modified CSO). Experimental results demonstrate that our proposed HYBRID algorithm outperforms peer research and benchmark algorithms (ACO, MPSO, CSO, RR and Exact algorithm based on branch-and-bound technique) in terms of efficient utilization of the cloud resources, improved reliability and reduced average response time.
77 citations
TL;DR: A light-weight application, CatchPhish which predicts the URL legitimacy without visiting the website, using hostname, full URL, Term Frequency-Inverse Document Frequency (TF-IDF) features and phish-hinted words from the suspicious URL for the classification using the Random forest classifier.
Abstract: There exists many anti-phishing techniques which use source code-based features and third party services to detect the phishing sites. These techniques have some limitations and one of them is that they fail to handle drive-by-downloads. They also use third-party services for the detection of phishing URLs which delay the classification process. Hence, in this paper, we propose a light-weight application, CatchPhish which predicts the URL legitimacy without visiting the website. The proposed technique uses hostname, full URL, Term Frequency-Inverse Document Frequency (TF-IDF) features and phish-hinted words from the suspicious URL for the classification using the Random forest classifier. The proposed model with only TF-IDF features on our dataset achieved an accuracy of 93.25%. Experiment with TF-IDF and hand-crafted features achieved a significant accuracy of 94.26% on our dataset and an accuracy of 98.25%, 97.49% on benchmark datasets which is much better than the existing baseline models.
70 citations
TL;DR: Results showed that in spite of low ECSA, PtxZn could not only have facilitated the single electron transfer to adsorbed CO2, but also showed better binding of the intermediate 〖CO_2〗^(•-) over its surface, and the lower bond energy between the mixed phase surface and -OCH3 compared to the phase pure catalysts has enabled higher CH3OH selectivity over Ptxzn.
Abstract: The electrochemical reduction of CO2 (CO2RR) to produce valuable synthetic fuel like CH3OH not only mitigates the accumulated greenhouse gas from the environment but is also a promising direction toward attenuating our continuous reliance on fossil fuels. However, CO2RR to yield CH3OH suffers because of large overpotential, competitive H2 evolution reaction (HER), and poor product selectivity. In this regard, intermetallic alloy catalysts open up a wide possibility of fine-tuning the electronic property and attain appropriate structures that facilitate selective CO2RR. Here, we report for the first time the CO2RR over carbon-supported PtZn nano-alloys and probed the crucial role of structures and interfaces as active sites. PtZn/C, Pt3Zn/C, and PtxZn/C (1 < x < 3) synthesized from the metal-organic framework material were characterized structurally and morphologically. The catalysts demonstrated structure dependency toward CH3OH selectivity, as the mixed-phase PtxZn/C outperformed the phase-pure PtZn/C and Pt3Zn/C. The structure-dependent reaction mechanism and the kinetics were elucidated over the synthesized catalysts with the help of detail experiments and associated density functional theory calculations. Results showed that in spite of low electrochemically active surface area, PtxZn could not only have facilitated the single electron transfer to adsorbed CO2 but also showed better binding of the intermediate CO2•- over its surface. Moreover, the lower bond energy between the mixed-phase surface and -OCH3 compared to the phase-pure catalysts has enabled higher CH3OH selectivity over PtxZn. This work opens a wide possibility of studying the role of interfaces between phase-pure nano-alloys toward CO2RR.
69 citations
TL;DR: In this article, polyphenylsulfone/multiwalled carbon nanotubes/polyvinylpyrrolidone/1-methyl-2-pyrroleidone mixed matrix ultrafiltration flat-sheet membranes were fabricated via phase inversion process to inspect the heavy metals separation efficacy from aqueous media.
69 citations
TL;DR: This work proposes an Intelligent Smart Energy Management Systems (ISEMS) to handle energy demand in a smart grid environment with deep penetration of renewables, and compares several prediction models for accurate forecasting of energy with hourly and day ahead planning.
68 citations
TL;DR: In this paper, the effect of secondary flow produced by V-ribs on the overall performance of a triangular solar air heater (SAH) duct was investigated using computational fluid dynamics (CFD) and exergy analysis.
TL;DR: In this paper, it was shown through first-principles density functional theory calculations that Bi and Zn doping introduces a resonance level in lead-free SnTe, and the dominance of the heavy hole valence band at room temperature in Bi-Zn co-doped SnTe leads to a record high room temperature ZT of ∼0.6 at 840 K.
Abstract: Lead free SnTe with a tunable electronic structure has become the front runner in eco-friendly thermoelectrics. Herein, we show through first-principles density functional theory calculations that Bi and Zn doping introduces a resonance level in SnTe. The dominance of the heavy hole valence band at room temperature in Bi–Zn co-doped SnTe leads to a record high room temperature ZT of ∼0.3 (at 300 K) for SnTe based materials. The increase in the Seebeck coefficient value due to the interaction between the resonance states and formation of the nanoprecipitates leading to an appreciably low lattice thermal conductivity of 0.68 W m−1 K−1 results in a peak ZT of ∼1.6 at 840 K. A record high ZTaverage of ∼0.86 with 300 K and 840 K as cold and hot ends, respectively, makes Bi–Zn co-doped SnTe a potential material for thermoelectric applications. This strategy of using two resonant dopants, to not only improve the room temperature ZT but also high temperature values, can very well be extended to other systems.
TL;DR: In this article, the effect of higher alcohol blends on performance and emission parameters of CRDI CI engine with various EGR rate was investigated, and it was concluded that up to 30% of the 1-pentanol can be used as an alternative to the diesel with a slight cost of performance.
TL;DR: This brief presents a novel seven-level (7L) inverter topology for grid-connected renewable applications that consists of ten active switches and one inner flying-capacitor unit forming a structure similar to conventional active neutral point clamped inverter.
Abstract: This brief presents a novel seven-level (7L) inverter topology for grid-connected renewable applications. It consists of ten active switches and one inner flying-capacitor unit forming a structure similar to conventional active neutral point clamped inverter. The proposed unique arrangement reduces the number of active, passive components and it does not require any sensor to balance the floating capacitor voltage, thereby reduces cost and complexity in the control system design. In addition, compared to major conventional 7L inverter topologies, the proposed topology is capable of boosting the input voltage by a factor of 1.5, thereby, eliminating the need for an intermediate boosting stage. In other words, it reduces the dc-link voltage requirement by 50%. To prove the advantage of the proposed topology over other recent topologies, a comparative study in terms of power components and cost is presented. The operation and performance of the proposed topology for various loading conditions are validated through experimental tests and measurements.
TL;DR: In this paper, the influence of nano-silica on hydration properties of binary, ternary and quaternary blended cement paste and mortar containing micro-to nano sized admixtures including fly ash (FA), ultrafine fly ash(UFFA) and nano silica in colloidal form (CNS).
TL;DR: It was noticeably depicted that 2/3rd of the KHSB groundwater quality falls under poor to very poor condition, and hardly 26% of groundwater available is portable, so this study contributes the effective use of multivariate statistics and WQI analysis for groundwater quality.
Abstract: Groundwater quality analysis has become essentially important in the present world scenario. In recent years, advanced technologies have replaced the traditional ones which are being helpful in simplifying the complex works. In this study, multivariate statistical analysis is carried out with the help of SPSS software for 45 groundwater samples of Kanavi Halla Sub-Basin (KHSB). The quality of groundwater is determined for various parameters which were analyzed and their concentration is correlated with other parameters using correlation matrix. The PCA technique is applied on water quality parameters, from which four components are extracted with 80.28% total variance. The extracted components suggest that the sources behind the higher loadings of each factor are by geological, agricultural, rainfall, domestic wastewater and industrial activities. Results of the Kaiser–Meyer–Olkin and Bartlett’s test conducted have value of 0.659 which is greater than the standard value (0.5). Based on water quality index (WQI), it was noticeably depicted that 2/3rd of the KHSB groundwater quality falls under poor to very poor condition, and hardly 26% of groundwater available is portable. Thus, this study contributes the effective use of multivariate statistics and WQI analysis for groundwater quality. It helps in understanding the hydro-geochemistry of the groundwater and also aids in minimizing the larger set of data into smaller set with effective interpretation.
TL;DR: Novel phishing URL detection models using Deep Neural Network, Long Short-Term Memory, and Convolution Neural Network are proposed using only 10 features of earlier work, which achieves an accuracy of 99.52% for DNN, 99.57% for LSTM and 99.43% for CNN.
Abstract: Phishing is a fraudulent practice and a form of cyber-attack designed and executed with the sole purpose of gathering sensitive information by masquerading the genuine websites Phishers fool users by replicating the original and genuine contents to reveal personal information such as security number, credit card number, password, etc There are many anti-phishing techniques such as blacklist- or whitelist-, heuristic-feature- and visual-similarity-based methods proposed as of today Modern browsers adapt to reduce the chances of users getting trapped into a vicious agenda, but still users fall as prey to phishers and end up revealing their secret information In a previous work, the authors proposed a machine learning approach based on heuristic features for phishing website detection and achieved an accuracy of 995% using 18 features In this paper, we have proposed novel phishing URL detection models using (a) Deep Neural Network (DNN), (b) Long Short-Term Memory (LSTM) and (c) Convolution Neural Network (CNN) using only 10 features of our earlier work The proposed technique achieves an accuracy of 9952% for DNN, 9957% for LSTM and 9943% for CNN The proposed techniques utilize only one third-party service feature, thus making it more robust to failure and increases the speed of phishing detection
TL;DR: In this article, a facile one pot solvothermal approach for the synthesis of V doped SrTiO3 nanocubes was employed for photocatalysis due to their tunable electronic structure.
TL;DR: A new boost inverter topology with nine level output voltage waveform using a single dc source and two switched capacitors, which eliminates the need for an input dc boost converter especially when the inverter is powered from a renewable source.
Abstract: This paper presents a new boost inverter topology with nine level output voltage waveform using a single dc source and two switched capacitors. The capacitor voltages are self-balancing and thus is devoid of any sensors and auxiliary circuitry. The output voltage is twice higher than the input voltage, which eliminates the need for an input dc boost converter especially when the inverter is powered from a renewable source. The merits of the proposed topology in terms of the number of devices and cost are highlighted by comparing the recent and conventional inverter topologies. In addition to this, the total voltage stress of the proposed topology is lower and have a maximum efficiency of 98.25%. The operation and dynamic performance of the proposed topology have been simulated using PLECS software and are validated using an experimental setup considering a different dynamic operation.
TL;DR: In this article, the authors reviewed the role of tilt angles and particularly locating the optimum tilt angle using different methods, which can be categorized mainly into mathematical model based, experimental based, simulation based, or combination of any of these.
Abstract: With the growing demand of economically feasible, clean, and renewable energy, the use of solar photovoltaic (PV) systems is increasing. The PV panel performance to generate electrical energy depends on many factors among which tilt angle is also a crucial one. Among hundreds of research work performed pertinent to solar PV panels performance, this work critically reviews the role of tilt angles and particularly locating the optimum tilt angle using different methods. The past data collected for analysis can be categorized mainly into mathematical model based, experimental based, simulation based, or combination of any of these. Single-axis tracking, dual-axis tracking, simple glass cover, hydrophobic glass cover, soiled glass, clean glass, partial shadow, use of phase-change material, computational fluid dynamic analysis, etc., are the novel methods found in the literature for analysis and locating the optimum tilt angle. For illustration purpose, few figures are provided in which the optimum tilt angle obtained on monthly, seasonally, and annual basis is shown. Research works are growing in the field of computations and simulations using online software and codes. Pure mathematical-based calculations are also reported but the trend is to combine this method with the simulation method. As the PV panel performance is found to be affected by number of parameters, their consideration in any single study is not reported. In future, work is required to carry out the experiment or simulation considering the effect of soiling, glass material, temperature, and surrounding ambience on the location of optimum tilt angle. As a whole, the optimum tilt angles reported for locations exactly on the equator line, i.e., 0° latitude, ranges between − 2.5° and 2.5°, for locations just above the equator line, i.e., latitude 2.6°–30° N ranges between 5° and 28°, for 40°–70° N, it is 29°–40°, and for 71°–90° N, it is 41°–45°. For locations at 2.6°–30° S, optimum tilt angles range between − 4° and − 32°, 30°–46° S, it is − 33° to − 36°, 47°–65° S, it is − 34° to − 50°, and for 66°–90° S it is − 51° to − 62°.
04 May 2020
TL;DR: This paper argues that existing approaches to comprehension do not adequately define comprehension; they are too unsystematic about what content is tested, and presents a detailed definition of comprehension—a “Template of Understanding”—for a widely useful class of texts, namely short narratives.
Abstract: Many tasks aim to measure machine reading comprehension (MRC), often focusing on question types presumed to be difficult. Rarely, however, do task designers start by considering what systems should in fact comprehend. In this paper we make two key contributions. First, we argue that existing approaches do not adequately define comprehension; they are too unsystematic about what content is tested. Second, we present a detailed definition of comprehension—a "Template of Understanding"—for a widely useful class of texts, namely short narratives. We then conduct an experiment that strongly suggests existing systems are not up to the task of narrative understanding as we define it.
TL;DR: It is observed that a combination of stylometric features and text-based word vector representations through ensemble methods can predict fake news with an accuracy of up to 95.49%.
Abstract: Social media is a platform to express one′s views and opinions freely and has made communication easier than it was before. This also opens up an opportunity for people to spread fake news intentionally. The ease of access to a variety of news sources on the web also brings the problem of people being exposed to fake news and possibly believing such news. This makes it important for us to detect and flag such content on social media. With the current rate of news generated on social media, it is difficult to differentiate between genuine news and hoaxes without knowing the source of the news. This paper discusses approaches to detection of fake news using only the features of the text of the news, without using any other related metadata. We observe that a combination of stylometric features and text-based word vector representations through ensemble methods can predict fake news with an accuracy of up to 95.49%.
TL;DR: In this article, an experimental study of 1-pentanol addition and EGR rates on the combustion, performance and emission of a CRDI diesel engine is carried out in this work.
TL;DR: A novel hybrid convolutional neural network (CNN) architecture for analyzing the students’ affective states in a classroom environment that predicts the overall affective state of the entire class.
Abstract: Predicting the students’ emotional and behavioral engagements using computer vision techniques is a challenging task. Though there are several state-of-the-art techniques for analyzing a student’s affective states in an e-learning environment (single person’s engagement detection in a single image frame), a very few works are available for analyzing the students’ affective states in a classroom environment (multiple people in a single image frame). Hence, in this paper, we propose a novel hybrid convolutional neural network (CNN) architecture for analyzing the students’ affective states in a classroom environment. This proposed architecture consists of two models, the first model (CNN-1) is designed to analyze the affective states of a single student in a single image frame and the second model (CNN-2) uses multiple students in a single image frame. Thus, our proposed hybrid architecture predicts the overall affective state of the entire class. The proposed architecture uses the students’ facial expressions, hand gestures and body postures for analyzing their affective states. Further, due to unavailability of standard datasets for the students’ affective state analysis, we created, annotated and tested on our dataset of over 8000 single face in a single image frame and 12000 multiple faces in a single image frame with three different affective states, namely: engaged, boredom and neutral. The experimental results demonstrate an accuracy of 86% and 70% for posed and spontaneous affective states of classroom data, respectively.
TL;DR: The proposed RS-coded MIMO UWOC system offers high reliability and power efficiency and it has the potential to be gainfully employed in IoUT applications.
TL;DR: In this paper, experiments were conducted in wide open throttle condition (WOT) for different speed ranging from 1200-1800rpm to 1800-rpm at an interval of 200 on a single-cylinder four-stroke variable compression ratio SI engine.
TL;DR: In this paper, first-principles density functional theory (DFT) computations were adopted to assess the potential application of a boron carbide (BC3) monolayer with point and topological defects as an anode mate.
Abstract: First-principles density functional theory (DFT) computations were adopted to assess the potential application of a boron carbide (BC3) monolayer with point and topological defects as an anode mate...
TL;DR: The proposed multiscale spatio-spectral feature based hybrid CNN model for hyperspectral image classification is compared against various state-of-the-art CNN based techniques and found to showcase a satisfactory result with less computational complexity.