The client’s objective was to extend the functionality of the distribution system through content personalization, providing end users with more flexibility in online bookings and content selection.
- The clients key requirements for the project were to serve real-time recommendations as well as the integration with service supplier systems and the formation of a multi-source database for travel tours and activities.
- The system was also created to generate recommendations based on users (recommendations are based on users with similar characteristics) or on items (recommendations are based on similar items, e.g. tours and travel activities).
- DataArt implemented a unique personalized recommendation engine, which offers new tours and activities based on a traveler’s personalized interests. The system was also created to generate recommendations based on users (recommendations are based on users with similar characteristics) or on items (recommendations are based on similar items, e.g. tours and travel activities).
In addition, DataArt delivered a data import solution, which aggregates travel tours and activities from different travel suppliers’ websites, such as City-Discovery, into a single database. All data is then combined in one format and presented to the end users in a curated and comprehensive format.
E-commerce website and mobile app
- Onsite Product Recommendations
- Onsite Content Personalization
- Segmentation & Insights
- Facebook & Instagram Ads
- Onsite Pop-ups
- Personalized Email
- In-Store Personalization
- Mobile App Personalization
- Product Recommendations on E-mails
- Adaptive Design
- Customising Recommendations
- Tracking Performance
- Advanced Analytic
- Tour and activities data that comes from partner sites sale logs is gathered in the backend
Data import solution & database management
- Targeted travel activities for users
- Extensive selection of travel offers
- Search across large volumes of data by different criteria
- Spark MLlib ALS-WR engine uses classification and a collaborative for processing the data and finding the recommendations.
- In-memory cached data model and MongoDB storage
Service supplier system integration
- Tour Managements System (Tour CMS)
- Tours and Activities Distribution Systems:
- Hotelbeds (Activity Content Integration)
Search functionality implementation
- Personalization AI engine
- Scalable real time search solution
- Support for multi-tenancy
- Track and micro-segment your visitors in real time
- Show each visitor the right set of recommended product
- Consistent experience across web, emails and mobile
- match visitors with similar tastes
Scoring mechanisms (Mohouts Rescorer)
- Geo location, spherical law of cosines to check product relevance
- Prejudice weather (eg. prefer in door locations because of weather conditions)
- Position in the product category tree