Listen Carefully to What Your Customers Say: Text Mining
Since the dawn of the computer age, understanding human language has been a goal of computer science research. In 1950, just after the first invention of the computer, Alan Turing proposed a test for computer intelligence. He introduced the imitation game, which was at that time merely a thought experiment.
In this game, a person asks questions via instant messaging or chatting to determine which of two hidden responders is a “man” and which is a “woman.” Two responders are in separate rooms from the questioner. One is a woman trying to help the questioner. The other is a machine, who is trying to confuse the questioner. Clearly, if the questioner identifies one of the responders as a computer, then the questioner knows the other is the woman, and the game is solved.
Turing, one of the greatest computer scientists of the twentieth century, introduced this machine in a paper called “Computing Machinery and Intelligence.” Although the philosophical aspects of the paper are interesting, it is just as notable for the assumption that computers would be able to readily engage in conversation.
More than six decades later, computers can do amazing things that Alan Turing probably never considered possible — from the ability to control machinery to the analysis of what were then unfathomable amounts of data to reasonable translations from one human language to another to winning Jeopardy. And yet, almost half a century ...