Intuitionistic Fuzzy Clustering to Information Retrieval from Cultural Databases

Content-based information retrieval (IR) involves low-level feature extraction and utilize similarity search methods applied either in the feature space or in derived higher-level semantic spaces. These methods assume that similarity is measured by accounting only the degree in which two entities are related, ignoring the hesitancy introduced by the degree in which they are unrelated. Aiming at semantically relevant IR from cultural databases, this paper proposes a novel intuitionistic fuzzy clustering scheme based on intuitive features and a similarity defined over higher-level patterns.