This paper proposes BARR, a framework for blog analysis. The framework considers blog content, author properties, reader properties, and the relationship between the author and the readers. Use of this framework can help identify hot topics, hot blogs, and opinion leaders. These identifications are useful in viral marketing, especially word-of-mouth marketing, to promote goods and services.

The evaluation is a reported comparison between the BARR framework and the TOPSIS method to identify hot blogs. The authors assert and show that the values of 11 parameters used to identify hot blogs were more evenly distributed in the application of the BARR framework than in the application of TOPSIS. The performance evaluation also shows that the weight ratio between quality and quantity of blogs can be decisive in opinion leader identification. The authors report further comparisons between the BARR framework, the Google search engine, and a blog Web site. They made all of these evaluations, together with the BARR derivation, after an initial ontology construction from an Apple iPhone search.

The BARR framework appears to be an initial formalized attempt at automated identification of hot blogs and opinion leaders. There is no call for further research; however, more actual examples of the application of the framework to other product and service areas by viral marketers would be interesting contributions to the evaluation of the BARR framework. This paper introduces a technique that we can research in other applications of social networking.