New Local Ranking Factor Coming Up?
He found that out via a paper, named “Hyper-Local, Direction-Based Ranking of Places“, written by four scientists from Google, Stanford University and Aarhus University (Note: Although I read the whole paper twice, there are very few things I understood about the way in which the algorithm would be changed as I am not a web/database analyst; maybe you would have better luck). The paper is scheduled to be presented at VLDB 2011 in Seattle.
The main problem of the study is the determination of the importance of points of interests or places in local search results, and how could the direction queries aid to solving this problem more accurately. They say:
A query that asks for directions from a location a to a location
b is taken to suggest that a user is interested in traveling to b
and thus is a vote that location b is interesting. Such user-generated
directions queries are particularly interesting because they are numerous
and contain precise locations.
The most important feature of the direction queries that is discussed in the paper is the frequency with which a place has been referred to. The four scientists write that this could serve as a signal as strong (or even stronger?) than user reviews. Immediately there a few questions that come to my mind:
– how about the people that do not drive (walk, go by bus/train)?
– how would that benefit places that are not points of interest to drive to?
– (as an user – Tom Health, has commented on the Greg Linden blog post) how would the cultural differences of using different means of transportation, rather than car, would be addressed?
– (and as the same user suggested) how would the issue be solved from the viewpoint of people who already know the place and don’t need checking the driving directions?
I guess we would have to sit and wait until September to find out the answers.