Not exactly sure how I found the website “I write like” but from the first moment I was mesmerized by the system. A simple idea in theory but very complicated in application. It’s a good example of where Artificial Intelligence (AI) has found a perfect application, to rapidly analyze language patterns, then make comparison of your writing style, to one of the more than 50 famous authors.
The other thing that’s very compelling about the “I write Like” website is the list of famous authors. Extra bonus is that you can learn about great new writers because the current list of 50 are very diverse, some of whom you likely will not have heard of, this will encourage more follow-up.
A great list of interesting writers, accompanied by thumb-nails of the Amazon linked books they’ve written. For example, my second and third random test of my writing (from this blog) returned the same result twice, that of a living legend, author, blog writer named Cory Doctorow.
I Write Like
|Dmitry Chestnykh, Coding Robots
|July 9, 2010
I Write Like is a website created by Russian software programmer Dmitry Chestnykh, founder of software company Coding Robots. The site analyzes users’ writing samples and, by looking for certain keywords, vocabulary, and style via a naive Bayes classifier returns the name of a popular writer the sample most closely resembles. It was launched on July 9, 2010 and, according to reports, has gone viral, getting over 100,000 visitors on July 13, 2010 and spreading quickly across other blogs and popular social-networking websites such as Facebook and Twitter.
According to the Toronto Star, the website’s popularity soared as a result of a series of rants made by actor Mel Gibson; New York City-based blog Gawker submitted transcripts of Gibson’s rants to find that the site website compared them to writings by Canadian writer and feminist Margaret Atwood. Film critic Roger Ebert tried the site and said on a tweet that “I Write Like thinks I write like Margaret Atwood, she writes like H. P. Lovecraft, and he writes like James Joyce“. The Star also reported the result that Ontario premier Dalton McGuinty‘s Children’s hospital in Ottawa celebrates big expansion closely resembles Joyce’s A Portrait of the Artist as a Young Man. The Guardian reported that Atwood herself tried out the website, and it said that her writing resembles that of novelist Stephen King in one attempt and like Joyce on another try. William Gibson also tried the site, which said his writing resembles that of Vladimir Nabokov. A transcript of a speech made by U.S. President Barack Obama in June 2010 has been compared to author David Foster Wallace, while the lyrics to Lady Gaga‘s song “Alejandro” have been compared to William Shakespeare. Other bloggers, including author Teresa Nielsen Hayden, have expressed anger and frustration to find that the website has compared their writings to that by The Da Vinci Code author Dan Brown. Despite the website’s early success, Chestnykh was cautious to say that its accuracy still needs improvement, as he has only been able to upload “a few books by some 50 different authors” into its database. He says that he intends to include advanced features such as “probability percentages” that a user-submitted passage resembles a certain author.
Below are the words copied from the About-page at “I write Like” official website:
I Write Like checks which famous writer you write like by analyzing your word choice and writing style and comparing them with those of the famous writers. Analyze your text
How does it work?
The algorithm pretty simple, and you can find it on every computer today. It’s a Bayesian classifier, which is widely used to fight spam on the Internet. Take for example the “Mark as spam” button in Gmail or Outlook. When you receive a message that you think is spam, you click this button, and the internal database gets trained to recognize future messages similar to this one as spam. This is basically how “I Write Like” works on my side: I feed it with “Frankenstein” and tell it, “This is Mary Shelley. Recognize works similar to this as Mary Shelley.” Of course, the algorithm is slightly different from the one used to detect spam, because it takes into account more stylistic features of the text, such as the number of words in sentences, the number of commas, semicolons, and whether the sentence is a direct speech or a quotation.
Do you want to learn more?
It’s open source!
We published our source code for everyone to review or reuse.
Is it correct?
It depends on your views on writing style. Certainly, you can’t rely on our analysis 100%. Try it and decide for yourself.