Wouldn’t it be cool to have an online travel guidebook that is based on the collective wisdom of hundreds of thousands of like minded travelers?

Well, researchers at the Department of Computer Science at Cornell University may have identified a way to do just that.  With the aid of a supercomputer, these scientists analyzed some 35 million Flickr photographs taken by more than 300,000 photographers from around the world and determined a unique way to automatically organize, label, and summarize them in such a way that someone could create an online travel planning guide that automatically suggests the ‘best’ travel sites, points of interests or landmarks one can visit during their vacation.

In their paper presented at the International World Wide Web Conference in Madrid this past April, which can be downloaded here, the Cornell scientists note that:

“In analyzing large photo collections, existing work has focused primarily on either structure, such as analyses of the social network ties between photographers, or content, such as studies in image tagging.  In contrast our goal is to investigate the interplay between the structure and content – using text tags and image features for content analysis and geospatial information for structural analysis.”

Put in non-scientific English, these guys developed some very complex algorithms that allows the multi-terabytes of Flickr travel photos and data to be organized much in the same way that the Web is organized…a method that is a real breakthrough with respect to mining information latent in very large sets of images.

Moreover, the Cornell researchers were able to mine this enormous mass of content and identified with surgical precision the world’s top travel destinations based on the number of photographs taken at each, and within each of those destinations, the top tourist points of interests, landmarks, etc.

Based on the number of Flickr photos, the top 5 cities on the ‘global traveler list’ are: New York City, London, San Francisco, Paris and Los Angeles (interesting to note that the US captured 3 of the top 5 spots, as well as 12 out of the top 25 cities worldwide!)

The top 5 Flickr most photographed points of interest/landmarks were: the Eiffel Tower (Paris), Trafalgar Square (London), Tate Modern Museum (London), Big Ben (London); and Notre Dame Cathedral (Paris)…curiously, note that no US point of interest/landmark made this list!

As you can see, the information that has been gleaned from Flickr can be leveraged to not only build an authoritative next-generation online source for recommending travel information about the most popular tourist destinations, points of interest and things to do, it also has the makings of a personal travel planning tool…bolting-on a booking box would, in our estimation, not be hard to do.

But the possibilities do not end there.  The study authors also noted at the end of their paper a very intriguing statement as to how their seminal work may be carried forward and what new and powerful information future research and analysis may yield:

“An interesting future direction is to relate {the geospatial data as a form of relational structure with content from the tag and image features] back to explicit relational structure in the social ties between photographers.  Preliminary investigation suggests that these can be quite strongly correlated – for example, we observe that if two users have taken a photo within 24 hours and 100km of each other, on at least five occasions and at five distinct geographic locations, there is a 59.8% chance that they are Flickr contacts.”

Wow…one can only imagine what additional ‘treasures’ can be mined through relational data sets like these; can’t wait for this phase of the research to start!

Although this study does not directly link to travel affiliate marketing, it does hold significant promise and impact for the travel industry in general, and specifically for the types of online tools, sources and influencers travelers will have at their disposal to help guide their future travel planning and purchasing decisions.

Towards that end, travel affiliates, travel suppliers and affiliate networks will by default be drawn into this process.  Travel Dividends suggest it’s important that all of us stay abreast of evolving technologies like this, so as to be in the ‘right place and at the right time’ and strategically deploy these technology tools in support of our business.

Are you a Flickrpreneur?

For some cool ways on embedding affiliate links in Flickr check out this 5StarAffiliate post.

We research Travel Affiliate Programs all day long.  Do you have a question you need answered?   Let us know…we’d love to hear from you.