An Enhanced Business Intelligence Approach for Increasing Customer Satisfaction Using Mining Techniques Yehia Helmy a , Ayman E. Khedr b , Shrief Kolief a, Eman Haggag a* a Faculty of Computers and Information, Information Systems Department, Helwan University, Egypt. b Faculty of Computers and Information Technology, Information Systems Department, Future University in Egypt *corresponding author E-mail:emanhaggag141278sams@gmail.com Abstract- The amazing growth of web creates a high demand for gaining knowledge about how the sites is being used. The extracted knowledge is the crystal source for sites’ administrators to take the enhancement decisions. Mining web data resources whether the users’ log files or users’ opinions is the main road to gain the required knowledge. Web Usage Mining (WUM) and Web Opinion Mining (OP) are the approaches for mining users’ logs and opinions. The paper proposes an adaptive frequency opinions’ positivity (FOP) algorithm based on linear regression algorithm. The algorithm used in investigating and measuring the mutual effect between WUM and OP. The experiment results were promising and supports greatly in improving the Business Intelligence (BI) level. Key words: Web Usage Mining, Web Opinion Mining, Business Intelligence, Regression Model, FOP algorithm. I. INTRODUCTION Web contains millions and millions of sites. To gain a unique characteristic for your site is a complicated process. This process needs to a precise knowledge in order to achieve its target. Mining deeply in web data resources is the road for winning the competitive knowledge in different platforms [18, 19, and 21]. Users’ logs and opinions are the vital sources for web data. WUM and OP are the mining approaches for gaining required knowledge (web knowledge) from these resources [7, 14, and 15]. WUM is the approach for mining in users’ log files while OP is for users’ opinions or reviews. Mining the users’ log files provides vital knowledge about how the site is being explored. This gives site’s administrator a comprehensive knowledge that helps greatly in improving the site’s structure. Otherwise, OP is seeking in users’ opinions. How the user finds the service or product, does service or product satisfy his needs, what is the user’s evaluation for site’s service or product and which are the features that are most important to users. These are examples for the gained OP knowledge. Studying this knowledge allows the decision maker to put his hand accurately on the strong and weak points in site’s services or products. From the previous, it could be seen that both of WUM knowledge (Kw) and OP knowledge (Kp) are vital knowledge to achieve the competitive advantage for any site in the world full of online business enterprises. Companies use BI to detect significant events and identify business trends to adapt quickly to changes in their environment [44, 24, 26]]. The implementation of BI is not an easy process because of data complexity [40]. Gained knowledge from mining web data (web knowledge) makes this process is clear and accurate. Web is the main and the most important way to communicate between customer and the enterprise. Web knowledge is the base stone for the decisions that affects the performance of the overall business. Gaining accurate and precise knowledge is the base of succeeding in achieving the BI [1, 20]. Most of previous studies focused their efforts in employing individually the gained web knowledge whether Kw or Kp to achieve the uniqueness and competitive advantages for the site. In this paper, researchers are interested in International Journal of Computer Science and Information Security (IJCSIS), Vol. 17, No. 4, April 2019 159 https://sites.google.com/site/ijcsis/ ISSN 1947-5500