Chin@mpton

Part 2: LBSN (Location Based Social Networking)

Abstract

Location-based Social Networks (LSNs) allow users to see where their friends are, to search location-tagged content within their social graph, and to meet others nearby. The recent availability of open mobile platforms, such as iPhones and Android phones, makes LBSN much more accessible to mobile users [16]. Besides, because our project is about location based discount information sharing, more accuratedly, it is a location based social networking in order to share hype-local discount information. Therefore, it is vital for us to analyse LSBN.

Introduction

There exist several differences between the traditional and the mobile environment. Complex (and expensive) billing models in the mobile context ask for short connection times and low data-volumes—requirements that do not exist in the flat rate dominated world of landline Internet connection. These monetary obstacles, together with the restricted input and output capabilities prevent the implementation of many of the currently successful features such as multimedia sharing or blogging on mobile devices. However, these devices exhibit other characteristics that are of advantage in mobile social networking. We believe that two key features are the user’s permanent reachability and location awareness. The location aspect is well reflected by the buzzword P3 (people-to-people-to-geographic-place) that recently emerged in the field [9].

Analysis of related work

Dodgeball is one of the first LBSN services that rely heavily on SMS to allow users to “check in” their current location and to find their friends and friends-of-friends within 10 block radius [10]. On the other hand, Loopt leverages GPS and other signal triangulation technologies to automatically sense device location, without requiring manual location updates.

Brightkite is a Denver-based startup, founded in 2005, that allows users to share their location, to post notes, and to upload photos through a number of interfaces, including Web, SMS, and Email. Recently, the company has also released a native client application on Apple iPhone and is planning a version for Google Android phones. These native client applications, like Loopt clients, leverage GPS and other on device technologies for automatic location sensing, though still requiring users to hit “check in” button to update location. Brightkite allows users to define their friends and subscribe to their activity streams, including locations they checked in, their posted notes, and their uploaded photos. A note is limited to contain maximum 140 characters, for users to share quick thoughts and short status updates. The “friendship” relation is mutual: a user X accepting Y’s friend request means that X and Y becomes each other’s friend [11]. A user may choose to protect her activity stream so only her friends can see her location/note/photo updates. A user may discover nearby people and browse their public activity streams. The posted notes and the photos are all tagged with user’s most recent checked-in location. Once a user checked in at a location, she is assumed to stay at that place until she explicitly checked in at another location. This mechanism gives users a complete control on when and where to share their location, addressing some privacy issues when sharing sensitive location information. When sharing user’s current location, Brightkite allows users to control the “granularity.” Namely, users can check in at a country, a city, or a zip code, without specifying the exact address.

FourSquare, most recently has become one of the most popular LBSN applications, launched in 2008 which engages its users in a game competition. User’s “check-in” at venues in order to be awarded points which contribute to their chart position. This fosters the engagement of users, which are encouraged to check-in as many times as possible. Initially, FourSquare allowed users to check-in only from about 100 cities in the US and in Europe; they have only recently removed this limitation [12]. Furthermore, the service enables the creation of bidirectional friendship links. The website audience has recently grown steadily, reaching about 100,000 members at the end of 2009. We used a public API to retrieve user friend lists and home locations with geographic coordinates. The duration of the crawl was 7 days from November 22 to November 28, 2009 and it was seeded with 1,000 randomly selected user identifiers. Due to a limit on the number of API requests that could be issued, we retrieved a subset of the entire network which contains information about 58,424 users [13].

LiveJournal is a community of bloggers with around 14 million active users as the end of 2009 [13]. Users can keep a blog or a journal and establish friendship connections among them. Each user provides a personal profile which often includes home location, personal interests and a list of other bloggers considered as friends. Friendship links may not be reciprocal. There is a public API to explore the social network, but it does not expose any method to get user profiles, where location information may be obtained. Thus, the crawling process involved both crawling the social network links through the API and downloading the HTML profile pages of the visited users. Seed users were acquired by accessing the public timeline over 24 hours and then 1,000 users were randomly selected among all the users retrieved. The duration of the crawl was 9 days, from November 2 to November 9, 2009, obtaining a sample of 1,502,684 users. Given the 1,226,412 users which provide location information, Salvatore Scellato et al. successfully obtained a meaningful geographic location for only 992,886 users [13].

Twitter is a social networking service which allows users to send short messages known as tweets. Tweets are composed only of text, with a strict limit of 140 characters: they are displayed on the author’s profile page and delivered to the author’s subscribers, who are also known as followers. Since its launch in 2006 it has gained a global and vast audience of millions of users all around the world [14]. Twitter does not enforce reciprocity in social connections: a user may follow another one even though the latter is not following back. Hence, the resulting graph is directed. Another key characteristic is the presence of a heterogeneous network structure, where a user may have many more followers than the number of users he/she is following, or vice versa. Twitter provides a public API to gather details on user profiles and follower lists. Due to a rate limit on API requests, it was not possible to collect information about all the Twitter users. The crawling process was seeded collecting 1,000 seed users from the public timeline, which shows a list of the 20 most recent tweets posted by users with unrestricted privacy settings to the entire service. The duration of the data crawling was 6 days from December 3 to December 8, 2009, gathering information about profiles and follower lists for 814,902 different users. Among them, 535,653 reported some information about their home location. Salvatore Scellato et al. have successfully geo-coded 409,093 users, translating their location information into a point on the Earth [13].

Table 1: Properties of the datasets: number of nodes N and edges K, average node degree <k>, average clustering coefficient <C>, average shortest path length <L>, average distance between nodes <Dij> [km], average link length <lij> [km], average node locality <NL> (in/out), average geographic clustering coefficient <GC> and reciprocity ρ[15].

There are many reasons to use LBSN. For example, the most straightforward reasons for neighborhood interaction are place-based [19]. That residents are collocated in a shared locale necessitates some form of governance that a corporate office or onsite management usually enacts. Thus, one purpose of communication and interaction is the exchange of information between residents and management about rental payments, utilities, repairs, noise, and other issues that directly relate to the shared space residents co-inhabit.

It seems that tying one’s personal social network to real world activities is proving to be extremely valuable. ABI Research predicts that location based services will generate about $2.6 billion this year in revenue and more than $14 billion in 2014. Currently, two-thirds of all Smartphone owners’ check in with a location based app at least once a week. It is expected that LBSN services will attract 82 million subscribers by 2013 [17].In Europe, the number of users of location based services is expected to increase from 50 million in 2008 to 130 million in 2014 [18].

References:

[9] V.A. Marco, et al. “VENETA: Serverless Friend-of-Friend Detection in Mobile Social Networking,” wimob, pp.184-189, 2008 IEEE International Conference on Wireless & Mobile Computing, Networking & Communication, 2008

[10] H. Lee “Mobile social networks and social practice: A case study of dodgeball”. Journal of Computer-Mediated Communication, 13(1):341–360, October 2007.

[11] L. Nan, C. Guanling, “Analysis of a Location-Based Social Network,” cse, vol. 4, pp.263-270, 2009 International Conference on Computational Science and Engineering, 2009

[12] FOURSQUARE. Foursquare. Everywhere. [Online] Available: http://bit.ly/coJPSY.

[13] S. Salvatore, et al. 2010. Distance matters: geo-social metrics for online social networks. In Proceedings of the 3rd conference on Online social networks (WOSN’10). USENIX Association, Berkeley, CA, USA, 8-8

[14] RJMETRICS, New Data on Twitter Users and Engagement, [Online]. Available: http://bit.ly/9JSNCf.

[15] GARLASCHELLI, D., AND LOFFREDO, M. I. Patterns of link reciprocity in directed networks. Phys. Rev. Lett. 93, 26 (2004).

[16]C. Guanling and R. Faruq. Analyzing privacy designs of mobile social networking applications. In Proceedings of the IEEE/IFIP International Symposium on Trust, Security and Privacy for Pervasive Applications (TSP), pages 83–88, Shanghai, China, December 2008.

[17] “82 million location-based mobile social networking subscriptions by 2013”. ABI Research Market Analysis Report, November 2008.

[18] Kate Shaw, (2010) “Comprehensive Guide to Location-Based Social Media”, [Online], Available: http://www.thesearchagents.com/2010/06/c…

[19] F. Marcus, “Facilitating Social Networking in Inner-City Neighborhoods”, Computer, Sep. 2006, volume 39, Issue 9, pp 45-50