Search

In the field of information security, phishing URLs detection and prevention has recently become egregious. For detecting, phishing attacks several anti-phishing systems have already been proposed by researchers. The performance of those systems can be affected due to the lack of proper selection of machine learning classifiers along with the types of feature sets. A details investigation on machine learning classifiers (KNN, DT, SVM, RF, ERT and GBT) along with three publicly available datasets with multidimensional feature sets have been presented on this paper. The performance of the classifiers has been evaluated by confusion matrix, precision, recall, F1-score, accuracy and misclassification rate. The best output obtained from Random Forest and Extremely Randomized Tree with dataset one and three (binary class feature set) of 97% and 98% accuracy accordingly. In multiclass feature set (dataset two), Gradient Boosting Tree provides highest performance with 92% accuracy.

Stock price forecasting, is one of the most significant financial complexities, since data are not reliable and noisy, impacting many factors. This article offers a machine learning model for the stock price prediction using Support Vector Machine-Regression (SVR) with two different kernels which are Radial Basis Function (RBF) and linear kernel. This study shows the Prediction and accuracy comparison between Support Vector Regression (SVR) and Linear Regression (LR) and also the accuracy comparison for different kernels of Support vector Regression (SVR). The model has used sum squared error (SSE) to determine the accuracy of each algorithm; which has shown significant improvement than the other studies. This analysis is conducted on the price data of about five years of Grameenphone listed on Dhaka Stock Exchange (DSE). The highest accuracy was found with Linear Regression model in every case with the highest accuracy of about 97.07% followed by SVR (Linear) model and SVR (radial basis function) model with the highest accuracy rate of about 97.06% and 96.82%. In some cases the accuracy of SVR (radial basis function) was higher than SVR (linear). But it was the Linear Regression which had the highest accuracy of all in every case.

Code reviewing becomes a more popular technique to find out early defects in source code. Nowadays practitioners are going for peer reviewing their codes by their co-developers to make the source code clean. Working on a distributed or dispersed team, code review is mandatory to check the patches to merge. Code reviewing can also be a form of validating functional and non-functional requirements. Sometimes reviewers do not put structured comments, which becomes a bottle neck to developers for solving the findings or suggestions commented by the reviewers. For making the code review participation more effective, structured and efficient review comments is mandatory. Mining the repositories of five commercialized projects, we have extracted 15223 review comments and labelled them. We have used 8 different machine learning and deep learning classifiers to train our model. Among those Stochastic Gradient Descent (SGD) technique achieves higher accuracy of 80.32%. This study will help the practitioners to build up structured and effective code review culture among global software developers.

Topics generated by topic models are typically reproduced as a list of words. To decrease the cognitional overhead of understanding these topics for end-users, we have proposed labeling topics with a noun phrase that summarizes its theme or idea. Using the WordNet lexical database as candidate labels, we estimate natural labeling for documents with words to select the most relevant labels for topics. Compared to WUP similarity topic labeling system, our methodology is simpler, more effective, and obtains better topic labels.

An increasing number of people are changing their way of thinking by reading news headlines. The interactivity and sincerity present in online news headlines are becoming influential to society. Apart from that, news websites build efficient policies to catch people’s awareness and attract their clicks. In that case, it is a must to identify the sentiment polarity of the news headlines for avoiding misconception. In this paper, we analyze 3383 news headlines generated by five major global newspapers during a minimum of four consecutive months. In order to identify the sentiment polarity (or sentiment orientation) of news headlines, we use 7 machine learning algorithms and compare those results to find the better ones. Among those Bernoulli Naive Bayes technique achieves higher accuracy than others. This study will help the public to make any decision based on news headlines by avoiding misconception against any leader or governance and will help to identify the most neutral newspaper or news blogs.

The most widely recognized relative directions are left, right, forward and backward. This paper has presented a computational technique for tracking location by learning relative directions between two intelligent agents, where two agents communicate with each other by radio signal and one intelligent agent helps another intelligent agent to find location. This proposed method represents an alternative approach to GSM (Global System for Mobile Communications) for the AI (Artificial Intelligence), where no network may not be available. Our research paper has proposed Relative Direction Based Location Tracking (RDBLT) model for understanding how one intelligent agent assists another intelligent agent to find out the location by learning and identifying relative directions. Moreover, three proficient algorithms have been developed for constructing our proposed model.