Published Paper Details:

With evolving technology, there are various ways and solutions for travelers to find information regarding their tour destinations. These include tourism websites, social network sites, blogs, forums, and search engines. Information found in these ways is not based on traveler�s preferences. Also, many places that deserve to be visited remain untouched due to the lack of public information. Many points of interest may be located within a range of dozens of kilometers from our homes, but usually we prefer to travel hundreds or even thousands of kilometers to visit some other well-known location. This paper proposes a web application which will help to address above challenges. In this paper, we address the development and the key features of a tourist guide application named Explore Mumbai. This application offers a set of searches and ?lters to facilitate the exploration of new locations, all of which is achieved using python framework- flask. MySQL is used to manage the database. Various web scraping techniques and tools such as Octoparse, request and lxml libraries are used for data collection and dataset creation. Features of places derived from descriptions, reviews of other people who visited those places and history of the user are the key features used to produce a list of recommended locations for the user. Other attributes such as location of the user is kept into mind when there is a cold start problem. To achieve the above goal KNN algorithm is used to compare latitude and longitude of various locations with current user location to recommend nearby locations. Content based recommendation is achieved using description of a hotel and then analysing similarities between other hotel�s description using TF-IDF. Collaborative filtering is used to compare other user�s review on a place with the user�s review. Cosine similarity is used to measure how many similar keywords are present in the descriptions and reviews