Scientists at Toshiba's Corporate Research and Development Center, in Japan have developed a system that offers shoppers advice on what to buy based on the product barcode and the current weblog buzz around the gadget. The team describes the system WOM Scouter this month in the International Journal of Metadata, Semantics and Ontologies.
If you've ever been at the mall hoping to choose, the latest mp3 player, camera, high-definition TV, you know how confusing the huge array of choice can be. You only have the sales assistant's word for whether a particular brand or model has a good reputation. Checking each and every model while in the store would be almost impossible, without a lot of mobile web browsing and manually entering model numbers into search engines.
Now, Takahiron Kawamura and colleagues have developed WOM (word-of-mouth) Scouter to allow shoppers to get the latest reviews for a product they are looking to buy simply while they are in store. The process involves taking a photo of the item's barcode with a cell phone camera. The WOM Scouter then looks up the item's meta data via the internet and gathers information from blogs and websites that review the product.
The WOM Scouter than uses natural language processing (NLP) techniques to analyze what blogs it collated are saying about the product and provides a straightforward positive or negative opinion on the product's reputation. Even the most confused shopper can make an informed decision on that basis, knowing that the blogosphere will support their choice.
The Toshiba team has tested WOM Scouter in a consumer electronics store and in a book store and suggest that the system represents a case of semantics used to provide an instant beneﬁt in a mobile computing environment. In essence, WOM Scouter is one of the first web 3.0 applications that utilises the fundamentals of the original web connectivity, the social media aspects of web 2.0, and provides a service based on the meaning, or semantics, of the data it handles.
WOM Scouter could be adapted not only to improve the shopping experience, but to help in choosing a movie to see, a restaurant at which to eat, or potentially whether or not to accept a job offer.
Source: Inderscience Publishers
Explore further: New algorithm identifies data subsets that will yield the most reliable predictions