Automatic Extraction and Generation of XML Documents from Financial Reports

Web services require XML formatted data. Human translation of business information from the rapidly expanding volume of documents to XML is labor-intensive and impractical. Computer programs can be built to extract domain-specific facts from web documents and convert them into an XML format. With a continual feed of web articles, such a system could be used to maintain an up-to-date XML knowledge base that could power web services for businesses. In this research, we build a system to automatically extract information from electronic international corporate financial reports, and translate this information into XML or XBRL (a well-known XML extension for accounting and financial data).