Software platform for bulk analysis of PDF documents

Tender description

The Department of Enterprise, Trade and Employment is seeking a suitably qualified and experienced ICT provider to provide a cloud-based system for the extraction of data from a large number (order of 3-5 million) of semi-structured annual report PDF documents held by the Department. These annual reports contain numerical data in tabular format as well as valuable qualitative information in the text fields of the documents. The software platform should be able to ingest the document files and be capable of recognising the various content items within the reports (tables, text fields, headings, etc.) which may be achieved through training the software on a training set of documents. The software platform should be able to locate, extract and classify numerical data from tables in the reports and arrange it into a large structured database of company financial data, which will be used by the Department for economic analysis and evidence-based policy research. The software platform should also make use of natural language processing (NLP) technology to locate and record specific qualitative information from the text fields of the reports and to extract semantic meaning from segments of text.