An Efficient Methodology for Resolving Uncertain Spatial References in Text Documents

Abstract

In recent decades, all the documents maintained by the industries are getting transformed into soft copies in either structured documents or as an e-copies. In text document processing, there is a number of ways available to extract the raw data. As the accuracy in finding the spatial data is crucial, this domain invites various research solutions that provide high accuracy. In this article, the Fuzzy Extraction, Resolving, and Clustering (FERC) architecture is proposed which uses fuzzy logic techniques to identify and cluster uncertain textual spatial reference. When the text corpus is queried with a spatial-keyword, FERC returns a set of relevant documents sorted in view of the fuzzy pertinence score. Any two documents may be compared in light of the spatial references that exist in them and their fuzzy similarity score is presented. This enables finding the degree to which the two documents speak about a specified location. The proposed architecture provides a better result set to the user, unlike a Boolean search where the document is either rated relevant or irrelevant.