Although there are plenty of movie recommendation sites on the Internet, none of them works the way NanoCrowd does. Movie recommendation sites work this way: you enter a movie you liked and the site will display a list of movies similar to the one you entered. Usually what happens is, the site maintains a huge database of movies with each movie properly tagged with multiple keywords based on the plot, actors, theme etc. When a user enters a movie name, the search engine retrieves the tags associated with that movie and then brings you names of other movies tagged with those same keywords. Sometimes that work but mostly it doesn’t and that’s why most movie recommendations sites are generally a waste. Sometimes the movies are rated by a few people behind the site to make the recommendation engine better. ClerkDogs is one such site that displays a list of handpicks movies and hence is a better recommendation engine than most.
NanoCrowd employs a different approach to rate movies. Instead of just one or a handful of reviewers, it collects hundreds of movie viewer’s comments from different sources like IMDB and Amazon, analyzes them and creates nanogenres for each movie. A nanogenre consist of three keywords that captures the essence of the movie and each movie can have multiple nanogenres.
When you ask us what movies are similar to a romance you like, not only do we tell you the movies most (and least) like your movie, we also recommend other movies that are: "magic, lovely, sweet" and separate them from the ones that are "tears, cried, emotions".
If your movie is a horror movie, we can recommend movies that are like it that are "undead, horror, '80s" movies, or some that are more "gruesome, gore, suspense", and even those that are "corny, cool, cheesy".
If you like crime movies, you might want us to separate those described as "sadistic, blood, torture" from the ones that are "sleuth, detective, mystery" or "noir, thriller, twists".
For instance nanogenres for Into the wild are “materialism, society, rebellion” and “wilderness, brave, breathtaking” and a few others, while those for Memento are “clues, intriguing, guessing” and “confusing, confusion, confused” (nice!)
Cliking on one of the nanogeneres brings more movies falling into the same nanogeneres. For each movie you search, NanoCrowd generates a “Movie in a nutshell” consisting of important keywords related to the movie so that you can quickly get a sense of the movie. Accompanying each movie there is a crowd rating and a crowd size implying how large the user comment base for that movie is. Hovering the mouse over a movie name pop-ups a brief description of the movie with links to read further reviews and critics comments.
NanoCrowd has another section called 2C List where you can save lists of movies you want to watch, movies you like, and more. NanoCrowd is also tied up with the online movie rental site Netflix, and if you have an account there you are able to add movies directly to your Netflix queue, and get recommend movies based on information already stored at Netflix.
NanoCrowd is quite unique and could probably lead to the rise of more movie recommendation sites in future that employ the same technique. Until that happens NanoCrowd is your best bet.
Comments
Post a Comment