Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Locality and Privacy in People-Nearby Applications


Published on

Presented at the Ubicomp conference, Zurich. People-Nearby applications are becoming a popular way for individuals to search for new social relations in their physical vicinity. This paper presents the results of a qualitative study, based on 25 interviews, examining how privacy and locality are managed in these applications. We describe how location is used as a grounding mechanism, providing a platform for honest and truthful signals in the challenging process of forming new social relations. We discuss our findings by suggesting theoretical frameworks that can be used to analyze the social space induced by the applications, as well as to inform the design of new technologies that foster the creation of new social ties.

Published in: Technology, News & Politics
  • Be the first to comment

  • Be the first to like this

Locality and Privacy in People-Nearby Applications

  1. 1. Eran Toch and Inbal Levi | @erant Locality and Privacy in People-Nearby Applications
  2. 2. Acknowledgments • Inbal Levi • Israel Science Foundation • Israel Ministry of Science Twitter: @erant
  3. 3. 3 Is the human race on its way to extinction?
  4. 4. 4 Because we spend all our time playing with our Ubicomp Gadgets?
  5. 5. 5 And do not meet people anymore?
  6. 6. 6 However, there is hope. There are examples of Ubicomp applications that help people meet strangers.
  7. 7. People-Nearby Applications (PNAs) People-Nearby Applications (PNAs) are used to meet strangers within a geographical proximity. 7 10km
  8. 8. and Blendr, Badoo, swiit locally, SupperKing, LeftoverSwap, highlight, Sonar, skout, grindr, speed flirt, sayHi, circle, banjo, foursquare, Twoo, meet me, Tinder, 4singles… Download estimates by; pictures from the AppStore. Grindr ~17M Downloads Highlight ~1M Downloads Skout ~31M Downloads
  9. 9. PNA as a Model 9 230K Downloads Lyft JobCompass ~720 Download Grub With Us PNAs are becoming a model for finding rides, jobs or even dinner invitations.
  10. 10. Pictures from the Highlight AppStore A Typical PNA 10 Finding users nearby Viewing a users’s profile Communicating through the inline chat
  11. 11. The Challenge • Creating new social relations is an action that combines high gain with high risk. • How users navigate between risk and opportunity? –How do users gain trust in one another? –How do users balance privacy and disclosure? –What is the role of location and locality? 11
  12. 12. 12 Background • Location disclosure to the social network (consolovo et al, 2005) • Location sharing in families (Boesen et al., 2010) • Location sharing in check-in services (Lindqvist, 2011) Location-Based Systems Computer-Mediated Communication • Language-based strategies to reduce uncertainty (Tidwell and Walther, 2002; Antheunis Et Al. 2011) • self-disclosure on dating sites (Gibbs et al., 2011)
  13. 13. Our Study A qualitative study, based on interviews with 25 users of Skout, SayHi and Blendr. 13
  14. 14. Recruitment • Participants were recruited by sending direct messages to users around Tel Aviv. • Overall, messages were sent to 320 users –20% of these participants responded –50% of those agreed to be interviewed and agreed to the institutional ethics consent form. 14
  15. 15. Participants • Gender –18 males and 7 female • Language –19 native Hebrew speakers and 6 native Arabic speakers. • Objectives –Chatting (14) –Dating and hookups (12) –Making new friends (6) 15
  16. 16. Methodology The semi-structured interviews were carried out using the inline chat system in Hebrew. 16 Interview Transcription Analysis Verification
  17. 17. Location
  18. 18. Interaction Radius • Participants interact with people who are 50km away from them, on average. • With a standard deviation of ±20km. 18 50km
  19. 19. So Close • 21 of the participants actively use location to decide who to interact with. • Most users were looking to interact with users that are in their vicinity: –”Of course that location is important. I want to find someone to meet, so why should I waste time on just talking? I won’t talk with someone who is 200 kilometer away.” (P2) 19
  20. 20. Far Away • However, 6 of the participants specifically search for users who are far away: –“I am interested in the Far East. I am here to meet people from Asia. ” (P3) –P12 was looking to talk with people from the U.S: ”I am here mainly to improve my English.” 20
  21. 21. Social Circles • We asked participants to characterize the people they interact with on PNAs. –15 participants: fundamentally different than the people they regularly interact with on Facebook. –6 participants: interact with people in a different language than their own. 21
  22. 22. PNAs in Urban Environments • A distance-based search in most urban areas covers heterogeneous neighborhoods. –A divergence from previous models of online social relation forming (e.g., Parks and Floyd, 1996) • So can locality lead to heterogeneity? 22
  23. 23. 23 Trust
  24. 24. Trust • Most participants expressed some negative perceptions of trust in other users. • 6 participants explicitly report some negative perceptions: –“Deception is very easy. Everybody can fill in fake details into their profiles.” (P10) –“Everybody wears a mask when they are online”. (P9) 24
  25. 25. Trust Process 25 Participants describe a two stage process, based on ‘trust signals’ within PNAs and in other applications. Established Communication 50% 25% 15% 10% Face-to-Face Meeting 50% Phone Skype Facebook Whatsapp Initial Contact 30% 30% 25% 15% Location Profile pictures Chat Signals Reporting & Blocking
  26. 26. Location and Trust • Location is used by 13 of the participants to establish trust in other users. –By grounding hypothetical knowledge: • “the location does tell me something about them, their culture, their perception.” (P21) • “This is a sane application, compared to what happens in others, where you really cannot know anything [about the other person]” (P8) 26
  27. 27. Distance and Trust • Proximity as familiarity –“Knowing the distance kind of helps me trust people. The feeling is that you can trust someone nearby as he or she are not total strangers.” (P7) • Distance as safety –“The further away I am, the more I believe there is no chance of meeting that person.” (P3). 27
  28. 28. Location as a “Hard” Signal • Location as an objective fact –“Trust is created... because they see where I am. If I am saying that I am somewhere near X and they see that its really so. (P20) • Followed by concerns of validity –“On the other hand, its also possible to use [location] for deception by blocking it on the mobile.” (P20) 28
  29. 29. Soft Signals • Profile elements and chat communications provide “softer” signals for trust: –“I try to ask a lot of questions, and to request photos to see if the person is trustworthy or not.” (P7) –“A profile picture and a picture album mean that the person is actually the person in the picture... Posts, vocabulary and location give an indication for intelligence and trust.” (P13). –“I ask some basic questions, and I continue to where it takes me. (P3) 29
  30. 30. 30 Privacy
  31. 31. Varied Privacy • Participants carefully manage their identity and their profile. • They vary considerable when it comes to identity: –14 had a recognizable picture. –6 had a non-recognizable picture. –5 did not have a photo of themselves at all. 31
  32. 32. Identity Framing • Users are wary of providing identifying details: –No users were identified using their real name. –“I never give my [phone] number here or my Facebook [page].” (P3). 32
  33. 33. Identity Markers • However, to establish trust and to reflect trust, identity is gradually revealed: –“I move them to Skype, with camera and all that.” (P8) –“Question: How do you make sure that someone is trustful? Answer: Facebook. You cannot fake pictures with friends, and friends’comments and stuff like that.” (P10) 33
  34. 34. The Application Gateway Waterfall 34 Established Communication 55% 25% 15% 10% Face-to-Face Meeting 50% Phone Skype Facebook Whatsapp
  35. 35. h,p://,e_soelter/6011656867 • Open environment: Multi- purpose and pseudo- anonymous. • Diverse, providing new opportunities unavailable in regulated spaces. • Grounding mechanisms are essential to establish trust. The Ecology of PNAs
  36. 36. Interaction Ecologies 36 If Facebook is suburbia PNAs are an urban street PNAs have some of the qualities defined in a public realm (see Loefland, The public realm: exploring the city’s quintessential social territory, 1998).
  37. 37. Violence • Violence is widespread, and described as a general phenomena by 10 participants. –4 females, out of 7, report on at least a single occasion of verbal violence. –3 males report on fraudulent attempts (1 case) and verbal harassment (2 cases). 37
  38. 38. Policing • The reporting mechanism is mentioned by 3 users as a crucial element of dealing with violence: –If I see that it [the chat] becomes harassing, I block right away.” (P3). –“If you discover that people are fake or perverts you just send the site’s team after them [report] and block them...” (P13). • Do users need a more secure ecology? 38
  39. 39. Wild PNAs “There are normal people here and there are less normal people here... but I think that its [the application] simplicity is what makes it nice. It’s somewhat primitive. Not like those fussy social networks. It’s primal.” (P8). 39
  40. 40. Take Home Messages • Location as a grounding mechanism. • The opportunities and pitfalls of open systems. • New frameworks are needed to understand the ecologies of open systems. –Theories from Urban Studies may be relevant to analyze Ubicomp ecologies. 40
  41. 41. Thanks! Acknowledgments • Inbal Levi! • Israel Science Foundation • Israel Ministry of Science Twitter: @erant