A Hybrid Approach to Discover MEC Interview Data With the Hierarchical Value Map of Social Networking Sites as an Example

As the booming of social network sites (SNSs), people adapt to communicate and share information via internet recently. According to great business opportunities emerging in SNSs, entrepreneurs strive to explore the potential needs inside users and then provide interesting feature functions on SNS platforms. The Means-End Chain (MECs) research method has been widely used to explore customers' perceived values in selecting products [1]. It is a good approach to help entrepreneurs finding the most appreciated product features. But however, while adopting MECs, researchers suffer the hassle of defining Attribute, Consequence and Value elements (ACV elements) from interview data. In addition, such context analyzing work heavily relies on researchers' subjective opinions, so that the research conclusions might be difficult to replicate and the contributions are limited. Therefore, this paper aims to propose hybrid miming techniques to automatically discover Attribute, Consequence and Value elements which are the most essential components in MEC approach. A case on studying customers' perceived values of social network cites is conducted by the proposed hybrid approach, and the experimental results show our method can discover the ACV elements effectively.