An interactive sentiment map generated by analyzing tens of thousands of hotel guest reviews has been posted online by Olset, a hotel data specialist, which scrapes the data from multiple sources, such as Booking.com, Hotels.com, Orbitz, and TripAdvisor.
The map, limited to the US, shows sentiments about more than 40 features mentioned in hotel reviews. It shows the most-positive and most-negative things from an aggregate state-level all the way to individual properties. For instance, in California people complain about staff much more than they do in Colorado.
Far more usefully, it slices and dices data about single hotels. Say you're a traveler who cares in particular about the air conditioning at a hotel. Olset's sentiment extraction and weighting system makes it intuitive to see how any given hotel's air conditioning has been evaluated by past guests.
In other words, the sentiment analysis divines more details and nuance than what is captured by reductionist one- through five-star ratings.
The map shows a slice of the San Francisco startup's global set of reviews, which it claims totals more than 50 million. The full data set with global content is reserved for its industry customers in 35 countries.
In most cases, those customers are OTAs looking for tools to better serve more relevant results to their travel searchers.
Olset also has customers in the corporate space who are using its sentiment data to present a traveler with more insight into any given hotel's reviews post-booking so that he or she gets a better idea of what to expect from their stay without having to read the reviews themselves.
Olset has nine full-time people on its team in San Francisco.
See the map, here, to judge for yourself if Big Data slicing and dicing can enhance the way reviews and ratings are presented to travelers: olset.com/sentiment-map
Earlier: Amadeus CheckMyTrip to add Olset hotel recommendations