Geo-Tagged Social Media Data as a Proxy for Urban Mobility

Twitter Connections Shaping New York City

“…They confirm the previously noticed pattern that communities of the networks of human mobility and interactions use to be spatially cohesive [1,[20][21]. They reveal key areas of NYC and are to a large extent consistent with the previous findings of [18], considering the network of locations across the city from a different perspective as described above. Reciprocal network actually provides a stronger similarity also capturing important features such as airports being attached to the core business area (Manhattan and Downtown Brooklyn).…”

“…This has been also validated on the city scale by using cell phone and taxi data [8]. Mobility patterns extracted from Twitter data have been successfully utilized to discover neighborhoods of New York City (NYC) [18]. This work considers the network of locations across the city from another perspective, being connected whenever a user residing in one location performs activity in the other and compares this network and its delineation results against the commuting network based on the Longitudinal Employment Household Dynamics Data from US Census and Twitter networks.…”

“…Cell phone connections [1, 5-6, 9, 19-22], credit card transactions [24][25][26], GPS readings [27] as well as various sensors data [12,28] serve as a useful proxy for human mobility, however its availability is limited largely due to the privacy concerns [2,4,11]. Social media, such as Twitter, on the other hand provides a broadly and more easily available alternative which once geo-tagged has been also proven useful for the human mobility studies [7,10,[17][18], despite its limited representativeness, being often seen as a challenge.…”

“…spatial network) will be then evaluated by applying it to the neighborhood delineation in New York City. While resulting neighborhoods seem meaningful, their similarity to the neighborhoods obtained from the user mobility network analysis [18] give rise to another research question -to what extent social connections and mobility are related?…”

Social Activity Networks Shaping St. Petersburg

“…However, the availability of this data is often limited, largely due to privacy concerns [15,2,16]. On the other hand, social networks, location services, and municipal services provide a broader and more accessible alternative and have been proven useful for human mobility studies [17,18,19,20,21]. The spatial projection of the social network structure is known to reflect useful geographical information.…”