The OIRE Method - Overview and Initial Validation
Huishi Yin, Dietmar Pfahl · 2018
Context: To lower the cost of Requirements Engineering (RE) and to help software organizations, in particular, small / start-up companies with small teams and little marketing and customer research budget, gain a better understanding of user satisfaction with existing products, we propose the Open Innovation in Requirements Engineering (OIRE) method to systematically analyze information that is openly available in the Internet. Goal: In this article, we describe the components of the OIRE method, explain what it can do and how it provides useful information to help software engineers and managers make better informed decisions. In addition, we present three typical use cases and a proof-of-concept of the method demonstrating the applicability of the method using real-world data collected from the Internet. Method: First, we choose app reviews from app stores and app monitoring platforms as the sources of input. Then, based on a list of features, we extract feature-related text as the input. Next, we manually apply the OIRE method and verify the accuracy of the outputs of each component of the OIRE method. Result: We found the following: a) The accuracy of component 1 and component 2 of the OIRE method exceeds 63% and the highest number is 88%; b) The overall accuracy of the OIRE method is between 50% and 61%; c) For the use cases chosen, the OIRE method has the potential to help software engineers and managers make better informed RE decisions.