As mentioned the SRS offers a set of predefined facilities the user can choose from based on Facebook categories (coffee shops, museums,.. etc) and SRS generates the route based on the number Likes for each location. One can assume that a higher pop-ularity leads to a higher vote, thus to more ‘gravity’ within the route generation. However, taking a closer look at the output shows that the system cannot differentiate in terms of popularity which is strongly related to the global range of the facility, if it is a ‘global player’ like ‘Starbucks’, a smaller local chain or a single business. That makes it difficult to differentiate if the recommendation follows local or global interests, if it is ‘insider (local) knowledge’ or the voice of the (global) crowd which is a highly interesting question.
The difference of the facilities of these categories is another challenge to the system. At the moment the system treats a shop same as a pub and a museum same as a restaurant, not taking into account the different types of customers it attracts (global / local) and their behavior. For further research it is very important to be able to differ-entiate between those two in order to understand the global and local relationship.
For a more general and global view on the subject it is important to point out that the environment used for the test represents an organically grown European urban development and does not take into account other city layouts. Urban layouts differ in terms of shape and scale, based on culture and history and have a big impact on peo-ple’s walking behavior .
(extract from my paper “Following the voice of the crowd: Exploring opportunities for using global voting data to enrich local urban context” as submitted for CAADfutures 2013).