Automatic Content Extraction on Semi-structured Documents
José Eduardo Bastos Dos Santos · 2011
Extracting specific content from certain types of documents can be a very challenging task, especially when developing a not so tailored solution and refraining from using explicit contextual information. In this paper, we address the problem of automatically extracting data from semi-structured documents through an unsupervised process based on an analysis of the document's own morphological composition. We also discuss how this approach can be applied to different types of documents, with special attention being paid to college transcripts. The success of our method is supported by extensive tests, from which we have drawn some authentic examples.