Business Tagging An application that helps a business know more about its customers.
Let us consider a restaurant as the targeted business The application requires bunch of images that were taken in the restaurant. More realistically, images can be obtained from social media sites since people usually tag the restaurant while uploading the photos. Once the application gets these images, it runs the images through Clarifia's API to obtain more detail about the image such as age group of people found, whether it attracts more of a lone customer or families or kids or couples etc. Whatever data Clarifai already provides, additional care is taken to accurately classify the image according to the application's requirements. With such clear and concise information about the customers, the business could proceed to advert their business in a better way targeting their majority customer group. One other application of such categorization is to provide a user a better idea of a business place. So it could be included as part of an app that suggests places according to user's preferences. So if a user wants to hangout with their friends or they wanna find some quite place to just relax and study or they wanna enjoy a meal with their family, the app is got all the information required about various places right there. The user could apply filters based on their preferences and
Business Review & Rating
To be more robust in ranking various business places, we factored in few more variables to calculate the rating besides the existing parameters such as users' experience and critics' reviews in text format. The additional variables considered are twitter feeds about a particular business place.
We took the details provided by Yelp as ground truth and added the information from twitter feeds. Basically, a sentiment analysis is performed on the comments to figure out whether it is a positive or negative comment and accordingly the rating is provided. Various other information considered are busy peak hours of a business place, how good a place provides the guaranteed utilities and any other information that the yelp data provides.
We also brought in the social ranking flavor so the rating has more meaning. But at the same time bounded the effect so that the system can not be abused by random users.We use the soacial media ( twitter ) as a agent to do that and Apache drill to query JSON data directly without storing huge data.
Log in or sign up for Devpost to join the conversation.