Attribute based classification and annotation of unstructured data in social networks

Abstract

Classification and annotation are two different and independent problems in social networks, but rarely considered together. Intuitively, annotations give evidence for the class label, which is used for classification and the class label gives evidence for annotations. Classification of unstructured data with annotation is complicated because of the low quality of the data and the rapid development of the social networks. Vast volumes of unstructured data are uploaded and shared in social networks, but only few of them are correctly annotated. This creates huge demand for automatic classification and annotation technique. The aim of the paper is to study attribute based learning for classification and annotation of the unstructured data. The work focuses on attribute based learning which is used to extract the features and to predict the attributes of the unstructured data. The method generates high level description that is phrased in terms of attributes which is used for the identification of unstructured data and this will improve annotation technique.