Principle Component Analysis and OCR algorithm for MNIST hand-written data

Companies always seek feedback from their customers about their products. Traditional questionnaires and surveys would be time and budget consuming. In the big data era with so many social medias and platforms like twitter, Facebook, yelp, etc, analyzing and visualizing the sentiment of people’s everyday comments regarding the products would be fastest and intuitive. We can not only get their attitude towards the product, but also learn the detailed problem they are facing with. We can further get an insight of how the negative reviews distribute based on customers’ geolocation. The ultimate goal of the project is using Twitter stream API, yelp API to get comments on the keywords/products the company cares about. Let's start with analyzing Tweets, where people talk about anything and everything. Tweets are narrowed down by keywords and hashtags. Pre-processing steps are used to eliminate noise of the text data. Then we do a sentiment analysis to achieve positive, negative and neutral statements. We’ll finally do a real time visualization on a map showing the distribution of the result.

Companies always seek feedback from their customers about their products. Traditional questionnaires and surveys would be time and budget consuming. In the big data era with so many social medias and platforms like twitter, Facebook, yelp, etc, analyzing and visualizing the sentiment of people’s everyday comments regarding the products would be fastest and intuitive. We can not only get their attitude towards the product, but also learn the detailed problem they are facing with. We can further get an insight of how the negative reviews distribute based on customers’ geolocation. The ultimate goal of the project is using Twitter stream API, yelp API to get comments on the keywords/products the company cares about. Let's start with analyzing Tweets, where people talk about anything and everything. Tweets are narrowed down by keywords and hashtags. Pre-processing steps are used to eliminate noise of the text data. Then we do a sentiment analysis to achieve positive, negative and neutral statements. We’ll finally do a real time visualization on a map showing the distribution of the result.