Page 370 - AI_Ver_3.0_class_11
P. 370

NLP is a subfield of Artificial Intelligence (AI) that enables computers to understand, generate, and enhance human
              speech. NLP has the capacity to query data using natural language text or voice. Most of us have used NLP. For example,
              NLP is at the heart of the technology, that powers virtual assistants like Oracle Digital Assistant (ODA), Siri, Cortana, and
              Alexa. NLP may be used on both written text and speech data.
              Some examples of NLP-powered tools are as follows:


                                            Intent Classification          Web Search






                                                            NLP-powered
                                    Auto Translate              tools             Email Spam Filtering







                                             Sentiment Analysis        Document Summary




                            Brainy Fact


                    The Georgetown-IBM experiment, conducted on January 7, 1954, was a landmark example of machine
                    translation.  Georgetown  University  and  IBM  collaborated  to  develop  the  project,  which  involved
                    complete automatic translation of more than sixty Russian sentences into English.





                      IBM Project Debater

              Can an AI system maintain an ongoing conversation and respond to what is said with each exchange? Moreover, can an
              AI defeat a human expert in an argument over a complex topic? Consider a well-known experiment to find out how AI
              works with spoken and written language.
              In 2012, IBM began developing Project Debater with the goal of
              creating a machine capable of more than just winning debates with
              humans. In other words, IBM Project Debater would listen to a series
              of competing arguments made by humans and respond to them
              intelligently, in addition to answering questions in human language.
              IBM’s goal was to create a system that could assist individuals in
              making evidence-based, bias-free decisions on complex problems,
              where the answers were not evident.

              A  winning  Project  Debate  could  be  prepared  and  presented  by
              following the given steps:

              1.  Understanding: AI has huge data at its disposal, several billion passages from articles, magazines, newspapers,
                  books, and journals. AI machine, on understanding the topic, create a vocabulary corpus and collate the data from
                  its whole pool of sources.

              2.  Reasoning: After the analysis, a short speech is made from different text collected from varied sources with every
                  minute detail presenting a compelling argument, in logical order, using good vocabulary.
                    368     Touchpad Artificial Intelligence (Ver. 3.0)-XI
   365   366   367   368   369   370   371   372   373   374   375