Recent Technological Advances in Natural Language Processing and Artificial Intelligence
This is an incremental review of existing technologies, relevant for researchers interested in NLP and AI progress.
The paper discusses IBM's DeepQA (Jeopardy) and WolframAlpha as examples of recent NLP/AI advances that tackle question-answering, highlighting their mechanisms and potential implications for achieving the Turing test by 2029.
A recent advance in computer technology has permitted scientists to implement and test algorithms that were known from quite some time (or not) but which were computationally expensive. Two such projects are IBM's Jeopardy as a part of its DeepQA project [1] and Wolfram's Wolframalpha[2]. Both these methods implement natural language processing (another goal of AI scientists) and try to answer questions as asked by the user. Though the goal of the two projects is similar, both of them have a different procedure at it's core. In the following sections, the mechanism and history of IBM's Jeopardy and Wolfram alpha has been explained followed by the implications of these projects in realizing Ray Kurzweil's [3] dream of passing the Turing test by 2029. A recipe of taking the above projects to a new level is also explained.