Cognitive Social Mining Analysis Using Data Mining Techniques

Abstract

Social media are based on computer-mediated technologies that smooth the progress of the creation and distribution of information, thoughts, idea, career benefits and other forms of expression via implicit communities and networks. The social network analysis (SNA) has emerged with the increasing popularity of social networking services like Facebook, Twitter, etc. Therefore, information about group cohesion, contribution in activities, and associations among subjects can be obtained from the analysis of the blogs. The analysis of the blogs required well-known knowledge discovery tools to help the administrator to discover participant collaborative activities or patterns with inferences to improve the learning and sharing process. Therefore, the goal of this chapter is to provide the data mining tools for information retrieval, statistical modelling and machine learning to employ data pre-processing, data analysis, and data interpretation processes to support the use of social network analysis (SNA) to improve the collaborative activities for better performance.

WITHDRAWN: Improving the Performance of Diagnosing Chronic obstructive Lung Disease Using Outlier Detection with Decision Tree Algorithm

“…Pradeep and Namrata compared the results of two feature selection strategies on seven different machine learning methodologies using data from thoracic surgery. For analyzing the performance of feature selection approaches, they used machine learning classi cation algorithms such as Nave Bayes, MLP, SMO, KNN, Linear SVM, CART, and RBF Network It demonstrates that after applying the over-sampling strategy to the dataset, all of the classi ers perform well in terms of performance measurements, but random forest surpasses the others with an accuracy of 83.6 percent [34]. Based on classi cation accuracy, the proposed study of this paper concluded that IQR with J48 improves the prediction of lung cancer patients' life expectancy after thoracic surgery.…”

Prediksi Diabetes Menggunakan Algoritma Naive Bayes dan Greedy Forward Selection

“…Pengklasifikasi Naive Bayes adalah pengklasifikasi probabilistik sederhana berdasarkan penerapan teorema Bayesian (dari statistik Bayesian) dengan asumsi penentuan nasib sendiri yang kuat (naif) [14]. Naïve Bayes dapat disebut juga dengan Simple Bayes dan Independence Bayes.…”

“…Pengklasifikasi Naive Bayes menganggap bahwa ada (atau tidak adanya) fitur (atribut) tertentu dari suatu kelas tidak terkait dengan ada (atau tidak adanya) fitur lain ketika variabel kelas diberikan [11]. Berikut persamaan Naïve Bayes [14]:…”

Prediksi Bedah Toraks Menggunakan Seleksi Fitur Forward Selection dan K-Nearest Neighbor

“…Penelitian pertama yaitu prediksi hidup dan mati pasien setelah 1 tahun melakukan bedah toraks [3] menggunakan Artificial Neural Network dengan hasil akurasi 90%. Pada penelitian selanjutnya, algoritma Naïve Bayes, J48 Decision Tree, PART (Partial Decision Tree), OneR (One Rule), Random Forest Tree, Decision Stump digunakan dengan hasil terbaik pada algoritma Random Forest Tree sebesar 95.65% [4] dan penelitian [5] yang menggunakan algoritma Multilayer Peceptron (MLP), J48 dan Naïve Bayes dengan hasil terbaik sebesar 82.3% pada algoritma MLP. Penelitian penggunaan algoritma KNN (K-Nearest Neighbor) pernah dilakukan untuk sejumlah dataset yang diantaranya pada penelitian [6] menggunakan dataset RSS (Really Simple Sindication) dengan hasil akurasi baik, penelitian [7] K-Nearest Neighbor termasuk kedalam sepuluh algoritma yang paling banyak digunakan untuk penelitian data mining [9].…”