Data Science



Travelling Recommendation

When we plan to travel there is always complexity on deciding the places, spending time, distance, etc. Picture a local travel guide who can understand you and provide the recommendations of a place like where to eat, what to visit, where to stay, hotspots of the locality where the user will be more interested, and all other particular details needed by the user similar to an itinerary. Travelling will be made simple. Our recommendation system does the same by understanding the user data and try to provide its best outcomes like the place, travel time and plans for a complete trip.

Client Profile

Our client is one of the leading travel agencies providing all the necessary support for their customer in travel needs. Overview The primary objective of the project is to identify the travel spots that are most suitable for users.

Key Challenges

The key challenge faced is identifying the neighbourhood and tourist attractions to formulate a traveling plan according to the users. Data collected for a user profile was very minimal and data of places or spots are sometimes structured or semi-structured. Prediction of the weather on a travel plan was yet another key challenge.

Our Approach

Our approach involves the collection of users & places data, finding the similarity between them and considering other factors like review and rating of various users to prioritize the places and finally applying the weather, distance filtering to find the most suitable place for the user.

The Solution

The data collected from different sources are structured according to our recommendation standard and stored inside the database. A classifier is created to identify the similarity of user preferences based on structured data. After getting the similarity data, we use the Google Map and WeatherAPI to filter the location according to environmental factors and provide the best recommendation.

Technology Used

  • Machine Learning
  • Programming Language: Python
  • Back-end: Combination of Hadoop and pyspark
  • Cloud: AWS ECS, S3 and RDS
  • ML Libraries: Pandas, NLTK, Gensim and SKlearn
  • API: Rest API like google Analyse and google map
  • Front-End: Django and Flask


The travelling plan created by the recommendation engine was good as we examine the results of the weather, cost-effectiveness and distance travelled compared to the manual planning.

Download Travelling Recommendation Case Study


Send download link to:

Other Case Studies