BookVibe is a three-year old Big Data venture that has developed a natural language search technology that allows users to search their tweets and extract the books that are being discussed by the people they follow on Twitter. The service is free to use for consumers and is looking to generate revenue by mining the data for book recommendations and book discovery and offering the relevant analytical information to marketers, publishers and authors.
Founded by Ramesh Haridas, BookVibe is a unit of Parakweet, the developer of the natural language processing technology that allows the company to crunch through hundreds of millions of tweets and detect and extract tweets with very specific meanings from normal language. In June Parakweet received $2 million in investment funding from Scott Banister (and early investor in Zappos and PayPal), Alan Braverman of Yammer and others.
In a phone interview with Haridas, he said the company’s natural language processing technology can search through about 500 million tweets a day and detect 100,000 tweets that are not only specifically about books, but whether the tweets deliver positive or negative comments about the book. He claims the technology has a 96% accuracy rate for identifying tweets about books.
The company is based in Palo Alto, has six fulltime employees—primarily computer and data scientists and engineers—and has several clients, including a national retailer and a large TV company although Hardas said NDAs prevented him from naming them.
“We started looking at the large amount of data on Twitter, trying to figure out how to get information out of it; using it for recommendations about food, travel, books or movies,” he said. Parakweet’s technology can identify “tweets and status updates that are very specific, tweets that offer opinions. We set out to capture this info for books and movies,” he said. Haridas says the technology can extract and isolate “the sentiments around a book; we can detect an intent to read a book as well as whether someone has already read a book.”
BookVibe also offers its users the ability to look at the book discussions on other Twitter users feeds. Under its Explore link, BookVibe can extract the book discussions on the feeds of celebrities, top authors, journalists and venture capitalists. In fact a BookVibe user can put any Twitter handle into the BookVide search engine and see what books that person’s followers are talking about.
Haridas said BookVibe offers consumers an easy channel for book discovery and offers publishers as well as the growing number of self-publishers, “tools for promotions and for analyzing their audience.” Haridas said, “publishers have no data about their readers, Amazon keeps it all, we can tell them who their readers are, what else they are reading and what devices they use.”
Haridas says the service is ideal for tracking book promotion—lower prices and track the results. BookVibe also rates the “social sentiment” around a book, rating each book as low, medium or high, based on “the overall social discussion around a book,” he said. “Publishers can use our tools to target Kindle owners and tell what books they have purchased.”