ApacheCon EU 2016

Although current big data systems for text processing can handle vast amount of textual data, they mostly perform keyword level or compositional semantic - level analyses, which limits their functionality and applications. We present an OpenNLP-based system leveraging a broad range of linguistic technologies for search, recommendation, content generation, sentiment, chat bots and social data analysis. We will learn how to leverage linguistic analysis features such as parts-of-speech, parse trees, parse tree-based relevance, discourse analysis, learning of discourse trees and tree kernel learning. We explore how to combine the best of both worlds: big data performance and linguistic style depth of analysis.