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

