Silicon Valley Code Camp : October 11th and 12th 2014

Alok Govil

unassigned
About Alok
Alok Govil is a full-stack technology architect and entrepreneur spanning both software and hardware with specific expertise in natural language processing, neural networks, graph databases, computer vision, analog and digital circuit design, electronic displays, MEMS, and end-to-end systems engineering. Starting programming at an early age of 11, he could program in machine code directly by 15 years of age. He is currently bootstrapping a startup in natural language processing and artificial intelligence, and also working with Qualcomm as lead technology architect in computer vision, machine learning and circuit design. Previously he was lead technology architect and project engineer for a multidisciplinary research program in electronic displays at Qualcomm. He has authored eight technical papers including two book chapters. He is recipient of a Gold Medal for being amongst top 25 students in India in National Standard Examination in Physics, 1994. He holds masters in electrical engineering from University of Southern California (USC).
{speaker.firstName} {speaker.lastName}

Speaking Sessions

  • Introduction to Natural Language Processing (NLP)

    10:45 AM Sunday   Room: 8338
    Natural Language Processing (NLP) and Understanding (NLU) aim to make machines process human languages like English. This session will provide a complete overview of the field from the basic structure of human languages to the state of the art. The session will focus on (a) deep theoretical understanding instead of mere use of pre-existing NLP libraries, (b) natural language understanding aspects instead of keywords-based analysis or text classification, (c) processing of 'English' 'text' and not speech recognition/synthesis or language translation. We will discuss NLP applications and challenges, language components, Chomsky’s hierarchy of grammars, parsing algorithms, word-sense disambiguation, logic and inference, language synthesis, and available test/training datasets.