There was a fascinating news item today on NPR about the use of text mining to understand something about a person’s private life. Ian Lancashire, Professor of English at the University of Toronto used text mining to study Agatha Christie’s novels. While he has done this with other authors before he came cross something particularly interesting in Agatha Christie’s 73rd novel.


It turned out that she was using a lot more of “indefinite words” (those that end in “thing” like something, anything) and used about 20% less of the vocabulary she normally used in other novels. Further research and discussions with other experts and anecdotes from her life led to the conclusion that she was suffering from Alzheimer’s at that point in her life even though she hadn’t been officially diagnosed.

The kicker is that the 73rd novel is called Elephants Can Remember and its central character is a female novelist losing her memory!

The NPR story talks about another interesting study called the Nun study about how early in life Alzheimer’s may manifest itself. Check it out.