Opinion mining (OM) is a process of extorting the opinions of customer about a service, product and policy. Recently, many research works have been designed for OM by using classification and ranking techniques. However, performance of existence techniques was not efficient for cross domain OM. In order to overcome such limitation, pairwise relative ranking based OM (PRR-OM) Technique is proposed. The PRR-OM Technique is designed with objective of improving the ranking efficiency (RE) for OM. Initially, PRR-OM Technique performs the preprocessing process in order to remove the unwanted data in customer reviews and thereby improving the OM performance. Next, PRR-OM Technique designs Pearson correlation based sentiment classification algorithm in order to classify the customer reviews into positive or negative opinions. This helps for improving the classification accuracy and reducing time for sentiment analysis. Finally, PRR-OM Technique used PRR in order to efficient mine the positive and negative opinion of customer about a particular product through ranking the word pairs based on their relativeness value. Therefore, PRR-OM Technique increases the RE. PRR-OM Technique conducts the experimental works on parameters namely preprocessing accuracy, classification time (CT) and RE. Simulation results confirm that the PRR-OM Technique is found to be better in terms of increasing the RE and reduction of CT as compared to existing works.