Automated content suggestion for document writing

Document writing can tend to be rather laborious and time-consuming. A good document requires substantial research to be performed on the subject of the document, followed by the additional task of aggregating the surveyed research into a comprehensive document with accurate and coherent content. In this paper, a dynamic content suggestion system is presented for document writing. The proposed system reduces human effort needed during the research phase by constantly analyzing the document context periodically and suggesting relevant content (text and media based), which can be added to the document by querying the web. The proposed solution helps in reducing the time to write documents by 42% as compared to the traditional methods of writing documents. Sophisticated statistical algorithms along with natural language processing technology are used to continuously determine the keywords and concepts from the content in the document. A web search is performed by ranking the most relevant keywords based on their score and the top five search results are chosen. The content from these web pages is then extracted and displayed on a tile-based user interface as a summary. This system aims to facilitate a faster and more efficient approach to document writing.