Twitter is a popular social networking and microblogging service that allows its users to send and read text messages, known as tweets. It was created in the year 2006 and since then, it has been generating about 340 million tweets and handling 1.6 billion search queries per day, through the company’s platform.
Many companies have tried to derive the information and value of those innumerable tweets. However, one cannot carryout a survey easily by analysing the twitter messages i.e., advising users about the best book to read or the best movie to watch based on tweets is not a simple task.
Natural language processing (NLP) is a combination of philological, computers and artificial intelligence. As NLP is concerned with the communication between computer systems and human natural languages, Parakweet uses this feature of NLP to derive the meaning from tweets, which are the form of human language input.
The company announced the launch of two products viz Bookvi.be and TrendFinder. The former is a free consumer-oriented book recommendation engine and the later is a social media dashboard for companies to monitor conversations on movies which is a paid product.
Its indeed a very tough task for the company, as there will be around 700,000 tweets a day discussing movies. Trying to figure out the best movie or a book based on these tweets based on text-matching techniques will yield around 40 million results as most of the books and movies have very common titles. Those results will not generate a proper output.
To avoid these issues, both the products use natural language processing to find out how common a title is on Twitter and also to figure out how a consumer is tweeting about a specific product. Based on these results, they make recommendations on the best readable book or a watchable movie.
In order to enable this task, Bookvi.be is built to recognize the words that are used in the tweet. For example, if some names a book suggesting that it’s a worst book ever, it would be insane for the product to suggest it just because it is mentioned in the tweet. So by identifying the kind of words used in the tweet, the product will consider whether or not to advise the book to the user.
These suggestions can be sent either to the users email on weekly basis as per their choice or they can get book suggestions depending on the people they follow on twitter by just typing in their twitter user name.
Though, there are many social recommendation tools, Parakweet tools are different as the accuracy is very high in them avoiding improper suggestions. They are incredibly precise and exact, reflecting the users opinions.
The best feature of the product is that it doesn’t require us to create a new social network or depend on friends for reviews on books or films. As most of the social networking users use them as a platform for expressing their views or interests, making content recommendations for things like books based on them may prove perfectly convincing.
Not just twitter, the company claims that they support Facebook and are about to add other platforms in near future. Products like this will help us to identify the most popular users choice.