Machine Learning Approach upon Text from Varied Publishing Formats

摘要

The paper aims toward reporting the approach to challenges for conversion of documents from varied publishing formats to machine readable formats. This research objective falls in the field of information retrieval. In order to represent the documents available in machine readable format from different publishing format, there is a need to identify, access, process and finally represent the information in such a manner which makes it ready for easy machine access. The above mentioned task involves different types of challenges as discussed in detail below. These challenges are mentioned with approach as proposed to solve information retrieval task in huge text corpora.