Building a Smart Travel Guide: Triposo's Suggestion Engine
summary
Triposo builds algorithmic travel guides that fit on a phone and work offline guides, covering 10,000 cities across ~180 countries with 3 million POIs and about 8 million app downloads. It crawls OpenStreetMap, Wikipedia, Wikitravel, Facebook and Booking, runs matching and opinion‑mining on combined feeds, and scores places for nearby, time‑based and weather‑based suggestions. Processing runs on two in‑office machines, pushes data via Dropbox/Google Sheets to S3, and uses Python with NLTK and immutable JSON logs for behavior signals.
AI-written summary · about
chapters (6)
- 0:00Introduction to Troso
- 0:44Algorithmic travel guides
- 2:47Nearby suggestions
- 6:33Weather-based suggestions
- 9:48Time-based suggestions
- 15:16Q&A session
related talks
28mMaking Recommendations without Data
28m"Where, When, Why, and How: Directi
28mBuilding a Recommendation Engine for Bus Search Based on Real-Time Feedback of Users
06mLearning from Hometown and Current City: Cross-city POI Recommendation via Interest Drift and ...
29mAn EV trip planner for Australia
01h 11mBuilding a Recommendation Engine with Neo4j and Python
38mLearning to Rank journey at GetYourGuide: Our Logbook #bbuzz
55mLeveraging Emerging Technologies in Travel for Competitive Advantage