Dimensionality reduction techniques for blog visualization

Exploratory data analysis often relies heavily on visual methods because of the power of the human eye to detect structures. For large, multidimensional data sets which cannot be easily visualized, the number of dimensions of the data can be reduced by applying dimensionality reduction techniques. This paper reviews current linear and nonlinear dimensionality reduction techniques in the context of data visualization. The dimensionality reduction techniques were used in our case study of business blogs. The superior techniques were able to discriminate the various categories of blogs quite accurately. To our knowledge, this is the first study using dimensionality reduction techniques for visualization of blogs. In summary, we have applied dimensionality reduction for visualization of real-world blog data, with potential applications in the ever-growing digital realm of social media.