Its been awhile since Wego launched a feature called “Guest Satisfaction”. If you have seen it and am wondering what is this all about, well basically we analyze hotel reviews from across the web and boil them down into a simple score (Guest Satisfaction).
How do we compute Guest Satisfaction Scores?
We examine hotel reviews and opinions from different sites around the web.
Semantic analysis of these reviews are done and by using sentiment analysis techniques, they’re grouped into Positive, Negative, and Neutral reviews.
Each hotel is then assigned a raw satisfaction score. This score takes into account the number of Positive/Negative/Neutral reviews. The raw scores are then normalized by computing their standard scores (also known as z-scores) within each destination. For example, all the hotels in Tokyo are compared to one another and their resulting scores normalized. The normalized satisfaction scores are then reduced to a scale of 5. This is the final satisfaction score that you see on Wego.
Why do we assign satisfaction scores within each destination instead of comparing hotels worldwide?
The reason is simple: assigning satisfaction scores by comparing hotels within each destination allows for a larger spread of scores. We want to make it easy for you to see which hotel is better reviewed within a destination.
If we’d compared hotels worldwide instead: most hotels in Wonderful Hotel City would end up with scores around 4 to 5. Most hotels in Rundown Hotel City would end up with scores around 0 to 2. This makes it really hard to compare hotels since the scores are so close. Imagine comparing hotels when they are mostly scored around 2.1-2.3!
Its been awhile since Wego launched a feature called “Guest Satisfaction”. If you have seen it and am wondering what is this all about, well basically we analyze hotel reviews from across the web and boil them down into a simple score (Guest Satisfaction).
How do we compute Guest Satisfaction Scores?
We examine hotel reviews and opinions from different sites around the web. Semantic analysis of these reviews are done and by using sentiment analysis techniques, they’re grouped into Positive, Negative, and Neutral reviews.
Each hotel is then assigned a raw satisfaction score. This score takes into account the number of Positive/Negative/Neutral reviews. The raw scores are then normalized by computing their standard scores (also known as z-scores) within each destination. For example, all the hotels in Tokyo are compared to one another and their resulting scores normalized. The normalized satisfaction scores are then reduced to a scale of 5. This is the final satisfaction score that you see on Wego.
Why do we assign satisfaction scores within each destination instead of comparing hotels worldwide?
The reason is simple: assigning satisfaction scores by comparing hotels within each destination allows for a larger spread of scores. We want to make it easy for you to see which hotel is better reviewed within a destination.
If we’d compared hotels worldwide instead: most hotels in Wonderful Hotel City would end up with scores around 4 to 5. Most hotels in Rundown Hotel City would end up with scores around 0 to 2. This makes it really hard to compare hotels since the scores are so close. Imagine comparing hotels when they are mostly scored around 2.1-2.3!
Take a look at our Guest Satisfaction at work
This entry was posted
on Tuesday, June 30th, 2009 at 9:49 am and is filed under Wego Labs.
You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.
