Market Basket Analysis (MBA) is an empirical data mining methodology to investigate the customer purchase behavior pattern in retail market. Association rules mining provides most extreme advantages to a retailer for product placement, making of promotions & discounts plans, customer segmentations and designing of different cross/up-selling strategies. Frequent itemsets mining is the first step for market basket analysis and mining different hidden selling patterns will boost the efficacy of marketing. Periodic behavior mining involves finding the set of items that periodically appear at regular intervals in transactional databases. Season-based pattern analysis will also be an effective basis for designing different market policies for retail businesses. Basically three famous measures support, confidence and lift are used to measure the quality of an association rule and the Apriori, FP-Growth and Eclat algorithms are most popular for mining the periodic and frequent itemsets. Customer Sentiment Analysis (CSA) plays a crucial role in understanding customer opinions and trends and helps to track the reputation metrics of products. Different Machine Learning based approaches are generally used to analyze customer sentiment faster and more accurately from customer feedback datasets. In Bangladesh, the super shop industry is growing rapidly each year. It is difficult to survive in competitive retail markets and facilitate sustainable growth without an effective business intelligence model to optimize the business strategies for long-term success in the retail landscape. In order to solve those issues, we propose an integrated approach to extract the significant insights (pattern/rules) by analyzing the customer transactional databases for designing new actionable business strategies in order to improve the customer satisfactions as well as increase the sales.