Avis MealHub-Richer Text Analysis

Project Proposal
The concept is derived from the social feature among humans that everyone likes to know information in a gist. There is always a speculation or degree of uncertainty about a restaurant to be visited on points like whether it would be good fit, whether it will be meet our expectations.
This project idea is to give the app user the experience of knowing about the place they are currently at before entering the place. Scattered information about a restaurant in google ,yelp consume lot of time to be analyzed and sometime it is difficult to find the correct review which gives you a better input about the place you chose to go grab your meal.
On top of the existing chaos, foursquare also gives information about trending restaurants in an area, which further influences your choices.
Our WebApp,helps the customer to find what the reviews from different customers actually infer by presenting them the relevant keywords associated with a review and their corresponding sentiments in presented in a well represented pattern to enable quicker understanding.

DataFlow
The web application provides the primary user interface with which the user can start the journey to use avis.Avis’s home page is powered by Google Map’s API and has ui design flow developed on bootstrap.js .

Please find below the step by step description of the process flow :
-The user input for the location preference is received at the home page.
-This location input is used to trigger a location based search for restaurants using Google Places API,Foursquare API, Yelp API.
-The static processed dataset available in Elastic search is queried based on the response for yelp api query. The static processed dataset is filtered using SPARK and passed through Alchemy API for review sentimental analysis. The tags obtained from the user based location search are used to implement relevant unique filter criteria to fine tune the result set.
-The Google Places API reviews are used as data source for kafka producer and Alchemy API uses the Kafka consumer to get its input reviews to be processed for sentimental analysis.
-The Foursquare API gives dynamic input about number of people at the restaurant adding to the trending attribute of the restaurant.
-The ranking algorithm is developed with restaurant‘s ranking from Yelp and its corresponding
-Listed restaurants sorted based on the ranking algorithm can be explored further to view the sentimental analysis about the reviews submitted.
-The review for a restaurant can be submitted through the review page and each review will be processed through Alchemy API for inclusion into further processing.

Technology Used:

Flask - Web server
AWS-EC2
KAFKA
SPARK -for data cleaning
Elastic Search -for data storage and indexing
Alchemy Language API
JQuery
Bootstrap API Calls

Data Source
Google Places API
Foursquare Venue API
Yelp API
Yelp academy dataset

Elastic Beanstalk Url:http://custom-env.cfiadyk93t.us-west-2.elasticbeanstalk.com/
Demo:https://www.youtube.com/watch?v=sGUHO9OMT6w