Extending datalog intelligence

Prominent sources of Big Data include technological and social trends, such as mobile computing, blogging, and social networking. The means to analyse such data are becoming more accessible with the development of business models like cloud computing, open-source and crowd sourcing. But that data have characteristics that pose challenges to traditional database systems. Due to the uncontrolled nature by which data is produced, much of it is free text, often in informal natural language, leading to computing environments with high levels of uncertainty and error. In this talk I will offer a vision of a database system that aims to facilitate the development of modern data-centric applications, by naturally unifying key functionalities of databases, text analytics, machine learning and artificial intelligence. I will also describe my past research towards pursuing the vision by extensions of Datalog — a well studied rule-based programming paradigm that features an inherent integration with the database, and has a robust declarative semantics. These extensions allow for incorporating information extraction from text, and for specifying statistical models by probabilistic programming.