Mobile social networking in theory and practice
First Monday

Mobile social networking in theory and practice by Giuseppe Lugano

Mobile social networks have gained the attention of the media, academy and mobile market. Despite of the robust tradition of network and community studies, mobile social networks are often addressed improperly. This paper presents a theoretical framework to study mobile social networking and examines the design implications of results of an exploratory study conducted with a group of 18 young adults in Finland. The findings of this study indicate that the quality of mobile applications offering social networking could be greatly increased by integrating the knowledge of two traditions that so far developed in parallel, namely the mathematical as well as social approaches to social networks.


1. Introduction
2. Mobile social networks
3. Methods and data
4. Results and discussion
5. Conclusions



1. Introduction

Technical advances of the Internet and mobile technologies have promoted new forms of social communication (Rheingold, 2002), allowing the maintenance of large distributed networks of contacts. Such tools can either complement or replace face–to–face meetings. Social software, such as e–mail and instant messaging (IM), already provide full support to interpersonal or group communication. Nevertheless, social networks remain often either invisible (e–mail) or private (buddy list). The introduction of online social networking sites (SNS) and mobile social software (MoSoSo) rendered them public, offering access to functionalities built around the interconnection of user profiles. Social connectivity is usually based on direct interaction or search for users according to specific criteria. However, social network data is not completely public: typically users can choose the level of visibility of their personal data through privacy management mechanisms. The first popular SNS, Friendster, was launched in 2002 and since then a number of similar applications have appeared and are now among the most popular Internet sites.

MoSoSo extends social networking to the mobile environment. Although MoSoSo applications are still relatively unknown, early versions were introduced at the turn of the millennium in parallel with the launch of smartphones, mobile devices with computer capabilities. Lugano [1] defined MoSoSo as

“a class of mobile applications whose scope is to support social interaction among interconnected individuals [...] exploiting the media convergence process and the increasing power of mobile devices to offer a variety of services.”

The technical convergence of computer and mobile networks opens opportunities for a synergy among SNS and MoSoSo, supporting social networking activity anytime and anywhere. At the moment, although much of the user–generated content (UGC) of SNS is created with mobile devices, the real potential of this synergy has not been realized. Despite of the diffusion of smartphones, few users benefit from mobile Internet services, as most of the mobile data traffic is still represented by Short Message Service (SMS). The under–use of advanced services is not only due to poor usability or marketing strategies, but also to the feeling that phone calls and text messages are the only essential social applications of mobile technologies. The mass adoption of MoSoSo applications does not represent only another means of generating revenue that could revitalize the mobile market, but also a chance for social change. Phone calls and text messages already have had a profound impact, providing the opportunity of being in perpetual contact (Katz and Aakhus, 2002) and breaking the traditional time and space boundaries. Ling (2004) observed that the most evident consequence of mobile communication can be found in the easier coordination of everyday life, but he also added that phone calls and text messaging have a poor scalability, as social coordination is efficient only within a small group. Smartmobs, or mobilization of the masses through mobile technologies (Rheingold, 2002), are neither interactive nor allow re–negotiation of time and place. In this context, MoSoSo represent a natural complement to the original mobile services extending from interpersonal to network interactions through the many–to–many communication paradigm. However, transforming SNS and MoSoSo from gadgets into tools for social change still requires an enormous effort. Among the most serious issues to include in multidisciplinary research, weneed to consider the impact of partial continuous attention (Stone, 2005), management of personal privacy (Raento and Oulasvirta, 2005) and issues of intellectual property rights (IPR) (Lessig, 2001). Technical solutions that underestimate the centrality of these issues could result in continued under–use of new services. Therefore, it is essential to acknowledge the human dimensions of technologies, by adopting a user–psychological perspective to understand needs and the significance of mobile contexts (Saariluoma, 2004).

This paper focuses on social context and mobile–mediated communication (MMC), an emerging area within the traditional perspective of computer–mediated communication (CMC) that emphasizes mobility issues. The convergence of computer and mobile networks is evident in the evolution of the mobile phone that was born as a communication medium to become a “portable multimedia computer.” The next step, which is already taking place, is related to its transformation into a “remote controller of people’s lives” (Lugano, 2007a). Although both the appearance and the components of modern mobile phones differ from old ones, many key elements of the user interface have not improved. For instance, the structure of the mobile phonebook has not been significantly improved since early versions. Mobile phonebooks do not integrate a user profile nor social networking features and mechanisms that support sharing, search and filtering of relevant data with other connected users. Although efforts in improving the design of mobile phonebooks have been done in the context of context–aware and ubiquitous computing (Oulasvirta, et al., 2005), overall interaction opportunities are not yet well supported by built–in features of mobile devices.

Design issues can be addressed from a conceptual perspective of mobile social networks (Lugano, et al., 2006), a socio–technical concept corresponding to patterns of interconnection with others emerging directly or indirectly through mobile communication. The concept follows the idea of social networks as personal communities (Wellman, 1988). The foundations of mobile social networking are centered on two building blocks: understanding social connectivity and exploiting it for interaction opportunities. The implementation requirement for these two phases is to collect and process personal and contextual data in real time in user’s devices. This paper addresses the first issue, understanding social connectivity in mobile social networks, discussing the findings of a user study in which empirical data was collected though a questionnaire that investigated aspects related to the attitude and actual use of the mobile phone. The qualitative aspect explored the user orientation towards mobile sharing, while the quantitative section was based on the analysis of mobile communication logs, containing one to two weeks of communication exchanges. Though mobile social networking may be linked to smartphones and mobile Internet applications, the dataset of the study consists of phone calls and text messages. The choice to focus on these two services is motivated by the fact that they are currently the only killer applications of mobile communication, heavily used by all users and available on all mobile devices. In addition, knowledge on patterns of social connectivity and orientation towards mobile sharing can be investigated in this context and applied to the design of more advanced social applications.

The rest of the paper is organized as follows: the theoretical background is presented in Section 2, which places the study of mobile social networks in the tradition of social network analysis (SNA). Section 3 contains the data and methods of the study, the results of which are introduced and discussed in Section 4. Conclusions and future extensions are illustrated in Section 5.



2. Mobile social networks

The study of mobile social networks can be placed in the context of SNA, a discipline dealing with the mathematical representation of patterns of social connectivity. The key concept of social networks was originally introduced by Barnes (1954), but here we adopt the definition stated in Wasserman and Faust [2]:

“a social network consists of a finite set or sets of actors and the relation or relations defined on them.”

According to Wasserman and Faust, “the presence of relational information is a critical and defining feature of a social network.” As several types of relations connecting individuals can be identified, social networks are inherently multidimensional.

Social networks became popular thanks to Milgram (1967), investigating the “Small World Problem.” In Milgram’s study, relational information was based on the forwarding of letters to acquaintances. In other words, two nodes A and B of the social network were connected if A either sent to or received a letter from B. The result of the study tells us that we live in a small world with “six degrees of separation.” Studies on complex networks contributed additional knowledge on the structural properties and dynamics of networks. In a well known study, Watts and Strogatz (1998) investigated the structure of the social world of Hollywood defining two actors as connected if they acted in the same film. Analysing the information present in the Internet Movie Database (, they concluded that the 225.000 actors were separated from each other by only four steps. Another important study provided an explanation to Milgram’s observation that most of the forwarding (i.e., connecting) was conducted by a small number of individuals with a connectivity significantly higher than the average. For Albert, et al. (1999) this is a property of a special group of complex networks that were called scale–free networks. The World Wide Web belongs to this category, as its highly heterogeneous distribution of links between the nodes of the network consists of most elements linked to a few others and a few network nodes, the “hubs”, which are far more connected than all the others. Hubs determine the small world effect by reducing the path length between any two nodes in the network. A study of Aida, et al. (2004) demonstrated that social networks of mobile phone calls present this characteristic. Mobile devices can be used as an additional research tool by social scientists to explore society, the most complex social network (Eagle and Pentland, 2006; Raento, et al., in press). González, et al. (2008) monitored, for six months, the movements of hundreds of thousands of individuals through their mobile traces with the goal of understanding the spread of viruses and diffusion of information. Their main result indicated that “human trajectories show a high degree of temporal and spatial regularity.” [3]

Candia, et al. (2007) used mobile social network data to investigate aspects of human dynamics and social interactions. Such rich data is not interesting only for mathematicians, but also for social scientists who can have deep insights to consumer habits, attitudes and routines. These findings are relevant for interaction designers as they provide concrete elements for enriching the user experience. After the technical convergence of computer and mobile networks, also the mathematical and social traditions in the study of social networks are converging. This trend is demonstrated also by a growing number of pioneering studies that are uncovering intriguing aspects of social structure and dynamics highlighting their design implications. For instance, in the Reality Mining project conducted at the Massachusetts Institute of Technology (MIT), Eagle and his colleagues collected over 350.000 hours of continuous data on human behaviour, including information about location, communication, proximity and activity, from 100 MIT students. They addressed research questions such as the evolution of social networks in time, the predictability of people’s lives and the way information flows. Such understanding was aimed at designing better tools for the coordination of interpersonal and group interactions.

Although advances in understanding the structure and dynamics of mobile social networks have been enormous, a conceptual framework to study this issue from the perspective of interaction design is still lacking. This paper is an attempt in this direction, regarding mobile social networks as they emerge from the communication of mobile phone users. More precisely, a mobile social network can be defined as the user’s patterns of interconnection with others emerging through the social use of mobile devices. Depending on the research question or goal, mathematicians, social scientists or interaction designers can define and explore one of the many existing types of social connectivity mediated by the mobile phone. For instance, Watts and Strogatz (1998) matched individuals according to their common profession, while the social ties studies by Milgram (1967) were linked to the actual communication exchanges. Other studies focus on the user location (González, et al., 2008) or on an aggregate of several variables (Eagle and Pentland, 2006). Despite of their different nature, all the variables of interest are part of the user profile. In SNS the data present in the user profile, mostly related to demographics and personal preferences, is used for connecting with other users. In mobile social networking applications, personal and contextual information is managed by social algorithms, procedures that collect and analyze large amounts of low–level data and return higher level knowledge to the user. A group of social algorithms could deal with aspects of social relationships, facilitating the establishment of new social ties or the maintenance of existing ones. Social algorithms are a generalization of current collaborative filtering systems, which are typically used in e–commerce platforms. In those cases, large amounts of data related to the past behavior of the user are correlated with the interactions of other users with the goal of presenting purchasing recommendations. In the context of MoSoSo, social algorithms could exploit well–known conceptual frameworks to enhance social interactions, for instance enabling connecting on the basis of shared interests. Social serendipity could be encouraged through the concept of homophily, originally introduced by Lazarsfeld and Merton (1954). Homophily is regarded as the degree to which individuals in a dyad are congruent or similar in certain attributes, such as demographic variables, beliefs and values. Homophily implies that distance in terms of social characteristics translates into network distance, the number of relationships through which a piece of information must travel to connect two individuals. The initial network studies explored homophily by demographic characteristics — such as age, sex, race/ethnicity and education (Bott, 1928; Loomis, 1946) — and by psychological attributes such as intelligence, attitudes, and aspirations (Richardson, 1940). In the mobile context, homophily relations could be investigated considering users’ similarity in mobile usage patterns. In a mobile social network based on homophily links, two nodes A and B are connected if they are similar on a selected attribute. Unlike in emotional closeness, homophily ties are unidirectional. Through the concept of homophily, interactions designers have a theoretical foundation to model interactions with strangers; network data can be used not only to establish new social ties, but also to enhance mobile interactions among already connected nodes.

In the mobile phone, the list of existing social ties is described in phonebook, which presents an overview of the mobile social network. A few studies investigated the importance of mobile phonebooks: Kuitto (2001), for example, used the information to assess the participants’ personal community. Her most relevant result deals with the significant overlapping between a user’s mobile social network and the real social network: only six of the participants reported that significant social contacts were not in the phonebook. A subsequent study compared paper and phone phonebooks and found that the knowledge present can contribute significantly to the understanding of the interaction between technology and sociability (Lonkila, 2004). The finite set of actors present in the mobile phonebook provides a useful account of the structural element of a mobile social network. Its relational aspect can be defined on the basis of data available on the mobile device, such as communication logs or sensor data. Therefore, the history of personal communications described by phone calls and text messages can be used to evaluate the properties of a social relationship, such as tie strength and reciprocity. Numerically, these values can be measured employing procedures similar to the ones described by Granovetter (1973) and Marsden (1990; 2005). For instance, “weak” and “strong” ties describing emotional closeness can be assessed through frequency and direction of mobile communication.

A final note on an alternative approach to MoSoSo design: mobile social networks are not the only conceptual perspective that can be used for the design of mobile social applications. An alternative approach is based on the consideration of the whole community and not of the individual as the basic social unit of a social system. Community–oriented social software supporting virtual communities already exists (Rheingold, 1993) and in the near future, also MoSoSo applications could be designed from this angle and lead to the emergence of self–organized mobile virtual communities.



3. Methods and data

3.1. Study setting

The dataset was collected during the autumn 2006 in Helsinki with the goal of gathering qualitative and quantitative information on mobile social networks. The former category corresponds to users’ general attitude towards mobile communication and in particular towards the sharing of digital information, while the latter is based on the actual mobile usage patterns and communication history. Participants were chosen on the basis of their acquaintance with the author for two main reasons. Firstly, it was easier to obtain access to personal information. Secondly, as friends tend to have a degree of overlapping acquaintances, it was possible not only to investigate the interviewer’s social network, but also to gain insight on other participants’ interconnections. For practical reasons, only the 20 closest local contacts were invited to a one–hour interview: 18 of them participated, seven males and 11 females. The majority of them (83 percent) were aged between 18 and 25 years. Fourteen (77 percent) were students and four full–time (22 percent) workers. Six students (33 percent) were also part–time workers; nine participants were Finnish, eight Italians and one French. Ten participants (56 percent) lived in their own apartment, five (28 percent) with friends, two (11 percent) with their family and only one (5 percent) with a partner.

3.2. The questionnaire

The instrument for data collection, a structured questionnaire, consisted of two main sections. The first group of questions, of a qualitative nature, was designed to obtain information on demographic and psychographic aspects of the respondents, while the second part of the questionnaire, of a quantitative nature, focused on behavioral variables connected to the communication with mobile social networks. Taken all together, the questionnaire contained 15 items.

The demographic variables included age, gender, nationality, contact information, working status and the number and type of relationship with prospective flat–mates. The psychographic attributes included attitude towards the utilization of mobile features and services, experience, and habits of mobile phone usage. The behavioral section corresponded to three tables filled respectively with information about phone calls, text, and multimedia messages stored in their mobile devices. More precisely, the variables recorded for each communication event were: contact phone number, communication channel (voice, SMS, MMS), direction of the communication (incoming, outgoing, missed), date and time. All variables were recorded directly from the mobile phone, for a total of 948 communication exchanges (67 percent calls and 33 percent sms). The multimedia messaging service (MMS) was not popular, with only one MMS found in the dataset. For each communication event, it was asked to label the communication partner with a social group (friend, family, partner, work colleague, acquaintance, other) and to specify the nature of the discussion. This variable consisted of two main categories, related to conversations of instrumental or expressive nature: the former group included six possible values (work, agreement, question, answer, location expression and recommendation) and the latter five (emotion, entertainment, chat, gossip and meta–comment). This did not include missed calls. Finally, it was asked to indicate the approximate location of the communication partner. The possible values were “the same city”, “a different city — the same country”, “different countries”.

Some complications in the process of gathering data arose because of technical limitations and human factors. Some participants did not have many text messages stored in their phone, even when they had reported sending many. The practice of deleting text messages was motivated by saving space or keeping just the meaningful ones. For this reason, the dataset included some old text messages, received even years before, typically of romantic nature. In one case the communication log was not complete, as the participant had removed the Subscriber Identification Module (SIM) a few days before; this removal action causes the reset of the communication logs. The lack of a uniform approach to storing communication logs added other complications: some old phone models did not have the ‘Sent’ folder, so it was not possible to retrieve information about sent messages. All devices recorded the name of the contact and the communication timestamp; however, phone models differ in the number of calls/messages kept in memory and in the way such information is stored. Old models usually store only the last ten incoming, outgoing and missed calls; furthermore, others keep track only of the last call with a contact, overwriting the timestamp of the previous communications. In other words, if A called B five times in a week, only A’s last call information was recorded in the mobile data logs.



4. Results and discussion

4.1. Experience and use of mobile phone features and services

The majority of the study participants (83 percent) reported having had more than five years of experience with mobile devices. Nokia was the most popular brand of mobile devices, as 94 percent of the interviewees owned a Nokia model. Old types of devices were still appreciated, as a third of the participants (33 percent) did not report interest in changing an outdated model to a new one. On the other hand, almost half (44 percent) had bought a new device in the previous year. Phone calls and the SMS were considered essential and accounted respectively for 33 percent and 67 percent of the communication. By contrast, the MMS service did not seem much appreciated, as only one MMS was received at the time of filling the questionnaire, representing a mere 0,001 percent of the mobile traffic of the study group. None of the participants sent an MMS during the study. This result indicates that the MMS service might probably never reach the success of the SMS. Smartphones accounted for 61 percent of the participants’ devices, but cameras were not regarded as an essential feature. Mobile Internet connectivity was considered important, but it was not popular: three people (17 percent) reported having occasionally browsed a Web site or read e–mail messages from their phone. The non–users of mobile Internet services motivated their choice by stating that it was “too complicated to use”, “too expensive” or that they felt “no need for it because they could connect to Internet from home”.

Some questions focused on users’ appreciation for contact and group management tools, essential features of mobile social networking. While early models represented a contact only with the pair <textual label, phone number>, more recent devices present more sophisticated records containing many details, which unlike the two basic fields are not mandatory. Our study revealed that users do not exploit the capability of newer phonebooks. When adding a new contact, all of the participants reported having inserted always only a contact name and phone number, while 22 percent of them had added sometimes also an e–mail address. More than half of the interviewees (56 percent) stated having linked a contact to more than one phone number.

4.2. Attitudes towards mobile sharing

Traditional mobile services, such as phone calls and text messages, are designed respectively as synchronous and asynchronous means for interpersonal communication. These services are suitable for social coordination of small groups of people, but they do not scale well for larger groups (Ling, 2004). The improved multimedia capabilities of newer mobile devices extend the scope of social interactions beyond the small social group. In practice, this opportunity depends on the sharing of personal and contextual data such as personal media, location and presence information. The disclosure of such information provides higher social awareness and greater flexibility in managing one’s mobile social network, but its highly sensitive nature raises privacy concerns that must be addressed in the design solution. Mobile sharing of personal resources can be regarded as the equivalent for the digital domain of self–disclosure in face–to–face situations. Studies on friendship highlight the relevance of the nature of the social relationship when investigating the issue of self–disclosure. A study from Newcomb and Bagwell (1995) suggests that friends disclose to one another more than do non–friends. According to Parker and Asher (1993), self–disclosure is related to positive aspects of the relationship, such as helping and companionship. Lugano and Saariluoma (2007) suggested modeling mobile sharing as a privacy–trust decision problem, an approach that takes into account the properties of a social relationship on the basis of contextual data stored in a mobile device. In order to decide whether a resource can be shared or not, the strength of the social tie is matched against the sensitivity of the resource, a value expressed at the initialization of the user profile. In the study, we investigated the influence of the nature of social relationship and resource in the user decision: the participants were asked to report with whom they would share some types of resources and how frequently (see Figure 1). The goal was to find out whether this question could be included in the steps to initialize the user profile and generate the default rule for privacy management described in (Lugano and Saariluoma, 2007). The resources included were: current location, presence (availability and away message), phonebook contacts, calendar events and personal notes, ring tones, mobile games and applications, and personal media (photos, videoclips, podcasts). Although the participants were not that confident with mobile sharing, they reported having had some experience with similar features presented by Internet services, such as IM and SNS.


Figure 1: Users' attitudes towards mobile sharing with friends/acquaintances (in percent)
Figure 1: Users’ attitudes towards mobile sharing with friends/acquaintances (in percent).


The findings do not contradict earlier results. Individuals were willing to share more types of resources and more often with their friends, but not with acquaintances or strangers. For instance, the most common answer concerning “sharing one’s location” was “sometimes” with friends and “seldom” with acquaintances. Many participants demanded the possibility of hiding their current location. Although SNS such as Facebook by default make it possible to explore even strangers’ list of contacts, the idea of sharing a mobile phonebook with others was not regarded as a good idea because of the personal nature of information. Considering presence, a feature aiming at a higher degree of social awareness, over one third of those (39 percent) interviewed would be ready to have it “always on” with their friends. Surprisingly, almost the same number of people (33 percent) would never share the same information with acquaintances. Similar results were found with the sharing of personal media: most of the participants would share them with their friends, but only 11 percent would show them to acquaintances. Concerning sharing of presence information and personal media, those who were against sharing explained their choice with “they are too private.” On the other hand, sharing ring tones, mobile games and applications was not considered dangerous to one’s privacy. The results show that 39 percent of the participants would like to always share ring tones while almost the same amount (33 percent) would never share them. A similar value was found for mobile games and applications, and we may suppose that the sharing feature would be used only by the people who already download ring tones and play mobile games, but it would not change the habits of others. Our findings suggest that the use of mobile phone features is affected by the nature of the relationship; therefore, its measurement should be taken into account in the design process.

4.3. Communication with social groups

Sharing of mobile resources could also take into account not only interpersonal relationships, but also participation in social groups. Mobile devices allow defining groups, but they do not offer a wide range of actions related to group management and communication. Elements of the mobile interface could be improved taking into account additional variables, such as the frequency of communication with social groups (see Figure 2). In this way, users could activate specific services only for the relevant groups. In our study, it was possible to analyse this variable with the data present in the mobile communication logs, as it was asked to the interviewees to specify for each event the social group of the communication partner.

Phone calls and text messages are used for regular communication with family, friends and partners, while business calls are less frequent. This low value might be influenced by the high number of face–to–face interactions at the workplace. Most users included in their mobile phone books also information service numbers, such as “Taxi” or “Hospital”. Such numbers were treated as a social group named “other”. Most participants (78 percent) reported having dialed service numbers, but not on a regular basis. Among them, the most popular service numbers were “Taxi”, “Answering machine” and “Mobile phone remaining credit”. Other services offered by companies are probably not well advertised, regarded as too expensive, or not considered useful.


Figure 2: Frequency of calls to groups (in percent
Figure 2: Frequency of calls to groups (in percent).


In her research on mobile use and social implications, Plant [4] found that:

“teenagers [...] often use their mobiles collectively, sharing information and showing each other messages, as well as comparing the frequency, the nature, and the variety of calls in rather competitive ways.”

Although not teenagers, the respondents demonstrated a similar behavior, as 94 percent of them reported having shown messages and multimedia content to friends and 44 percent having shown them to their family. This kind of knowledge could be used for instance to build multiple reputation systems, each one linked with the specific roles the user has in a social group.

4.4. Most popular times of mobile use

The analysis of the mobile dataset revealed that most communication occurred between 1 and 2 p.m. and between 6 and 9 p.m. (see Figure 3). No calls or messages were placed or sent between 6 and 8 a.m. Individually, people often placed calls in the same hours during weekdays, but with different patterns during the weekend. Communication activity was low during working hours (8–12 a.m. and 2–6 p.m.). The participants reported having placed calls or sent messages when ending an activity or when they were about to move to another location. This result is consistent with previous findings about the phenomenon of saturation of time (Rheingold, 2002; Ling, 2004; Castells, et al., 2006). Mobile devices record the duration of the last call, and this value was also asked to be reported in the questionnaire. Most calls (61 percent) were less than one minute long, and only 11 percent lasted more than five minutes. The average duration of the last call of the study participants was 1 minute and 51 seconds.

Time statistics could be used for personalized billing recommendations and optimization of phone contracts, enhancing the existing service provided by mobile operators that include statistics and usage fees in their invoices. In addition, temporal patterns of phone use could be also employed to enhance users’ mutual awareness and suggest the most suitable means for communicating at a certain time of the day. Temporal patterns were also investigated by Raento and Oulasvirta (2005) in the ContextPhone, which included a “fading hand” indicating the time passed since the user’s last interaction with the device.


Figure 3: Distribution of communication per hour of the day
Figure 3: Distribution of communication per hour of the day.


4.5. Spatial distribution of mobile contacts

Location–based Mobile Social Software (LBS–MoSoSo) is one of the most promising areas of development of MoSoSo that includes, among others, friend–finder applications. Sensor connectivity offers an additional communication channel for proximity interactions, providing users with freedom of choice among multiple solutions. Although the higher mobility of users produce also more distributed social networks, most interactions happen at a local level: almost 64 percent of the calls and messages were directed to or received from a person living in the same city, while six percent of the communication was directed towards other Finnish cities and the remaining 30 percent to other countries. The high amount of international communication is probably due to the foreign nationality of many of the respondents. The global scope of mobile communication is demonstrated by 16 different countries present in the dataset. Interestingly, 15 (94 percent) of them were European and only one from another continent (U.S.). Communication between Italy and Finland was particularly intense, as 17 percent (59 percent calls, 41 percent sms) of the interactions involved an Italian communication partner, often a family member. The comparison of communication patterns of Finns and Italians with their families suggests that cultural aspects play an important role: the former group places a relatively low number of cheap national calls, while the latter a large number of expensive international calls. In accordance to privacy management policies, regular updates of the user location could be exchanged by mobile devices, giving the possibility to the user to filter the entries in the phone book according to their location and suggesting the most suitable type of network connectivity.

4.6. Instrumental and expressive needs

The study participants were asked to label each communication exchange in their dataset with one label indicating the content or reason of the discussion. Missed calls, which represent a significant percentage (25 percent) of the total number of calls, were not labeled. Altogether, 791 labeled communication exchanges in two categories remained to be analyzed. The main finding confirms known results (Ling, 2004) on phone calls and text messages as flexible tools for serving both instrumental and expressive communication motives, with the former (56 percent) more frequently indicated as a motive than the latter (44 percent). In many cases people rely on others as sources of information, as almost half (49 percent) of the instrumental calls and messages were “questions”. Surprisingly, only 13 percent were labeled as “answers”. This finding is probably influenced by the requirement of labeling a communication with only one tag; in most cases, the conversation did not contain only an “answer”, but also information of other nature that was regarded as the main topic of the communication. The popular use of the mobile phone as a tool for social coordination is demonstrated by a large number of “agreement” conversations (22 percent) and “location” expressions (five percent). Discussions related to “work” matters were not popular (six percent). The expressive need of keeping in touch with others with short “chats” and “gossips” reported the highest value in its category (51 percent), followed by “emotion” interactions (34 percent). An important design implication of specifying instrumental or expressive tags for each communication concerns the possibility of generating tagclouds. Metadata could automatically be added by agents analyzing the content of the conversation or manually inserted by users. Both approaches should be included in the design, providing the user with choice in the tradeoff between precision and speed. Tagclouds could point to specific resources, places or needs to people, updating the current context by showing a visual summary of past interactions.

4.7. Size of the mobile social network

An important insight on the mobile social network of the study participants comes from the comparison of the size of mobile phone books and communication logs. Phone books typically (in 50 percent of the cases) contain between 100 and 200 contacts, in some cases (33 percent) they contain fewer than 100 and in some others (17 percent) more than 200. Communication logs revealed that people interacted via voice communication or text messaging in average with 21 different contacts. On average people communicated only with 24 percent of the people present in their phone book, following the “80/20 principle, or vital few — trivial many” stated by Pareto (1935).

The participants with the largest social networks were also the most socially active, engaging in interactions with 30 other individuals. The search for the hubs of the mobile social network has to possibly start from such persons. There were also some common acquaintances, on average three, among the participants of the study. The most connected persons, with one exception, were also the ones with the highest number of common acquaintances. The exception was represented by the authors’ partner, who had a limited social network, but an exceptional number of common acquaintances. In addition, this finding confirms that close contacts tend to share more and more common acquaintances with developments of romantic relationships (Parks, 2007). The key position of the author and the high level of social activity produced benefits also to the closest contacts, as they received network opportunities because of their strategic position and without the need of being very active. The relatively low overlapping indicates that friendship networks are becoming more and more distributed and sparsely knit (Wellman, 2001), with most people participating in several small groups, which are interconnected by a few key individuals more socially active than others.

The qualitative and quantitative sections of the questionnaire allow also comparing cognitive and real social ties measured respectively by perceived and actual frequency of communication with the mobile social network. The correspondence between these two measures can be evaluated by comparing the answers of the participants and mobile communication logs. Almost half of the participants (44 percent) stated having called their friends every day and slightly more of them (50 percent) once a week. The available communication logs cover between one and two weeks of actual communication. As the author was connected to all study participants either through a “strong” or “weak” tie, it was expected to find at least one call or message sent to him in 94 percent of the participants’ logs. However, the analysis of the communication logs revealed that only 61 percent communicated with him (22 percent via phone calls and 39 percent via text messages). This finding suggests the actual number of interactions is lower than the perceived ones, confirming that the quantitative measure of communication provides relevant information for the modeling of the mobile social network.



5. Conclusions

The findings of this exploratory study on mobile social networks suggest that communication logs, as parameters useful to assess properties of the social context, represent valuable knowledge for enhancing the design of mobile interfaces and personalized services. Three building blocks of mobile social networking applications have been recognized: user profile, mobile social network and social algorithms. The former two features represent the structural element, while the latter the class of procedures managing social dynamics. The necessity of integrating such elements in the design is going to be even more important in the context of ubiquitous applications processing an increasing amount of contextual data. Multidisciplinary research efforts are required in the area of user and network modeling, with a special attention to the definition of social algorithms, procedures that require the convergence of the mathematical and social approaches on networks.

From the perspective of design, the enhancement of the mobile phone book should be regarded as the most urgent enhancement, considering its importance and the current limitations. The findings of out study indicate that the new features added so far have not been significant, being perceived by users as “gadgets” rather than “added value”. The phone book is the ideal place where to embed the notions of user and mobile social networks, which can be regarded as a collection of user profiles. The user profile could consist of several sections, such as demographic variables (identity), psychographic attributes (interests, attitudes, opinions) and behavioural data (communication logs). Quite surprisingly, mobile social networking applications aim at integrating social networking without fully acknowledging the importance of the user profile. Social algorithms would make use of such information matching people on the basis of their similarity on a specific parameter (homophily) and infer aspects of social relationships, such as emotional closeness and reciprocity. At individual level, statistics based on communication or behavioral data can be employed to deliver relevant knowledge to the user by reducing the complexity of contextual information. Knowledge of the most frequently called or “nearby” contacts could be used to offer customised phone book views.

Additional research is also needed to understand the cognitive and social implications of the shift from storing knowledge and resources in digital memories rather than in human ones. Network resources can be accessed easily and quickly, even without explicit communication, thanks to the existence of a social connection. For instance, being connected in Facebook enables a user to receive automatically network updates from all connected contacts. The disclosure of personal resources contained in a user profile such as location, photos, current activity or mood transform them into social resources, which represent social capital if used for a purposive action by any of the social actors belonging to the personal community (Lin, 2001). Automatic access to network resources without the need for explicit communication can be regarded as an advantage, but it has also the potential to render weak ties even weaker and to shift the meaning of community further away from its original communitarian orientation and towards a mere support of an egocentric lifestyle (Ling, 2004). However, the sharing of resources and the real–time coordination of social action can also have a key role in the empowerment of decentralized self–organizing networks, whose contribution is important not only for the collaborative approach to problem–solving and innovation, but especially for the possibility of activating a network–based civil society (Viherä, 1999).

So far, the potential of peer–to–peer (p2p) networks has been regarded mostly from the perspective of computing, but it could be considered also for tackling the global challenges of the information society. Mobile communication has already produced a number of relevant social consequences: among them, the fading boundaries between public and private sphere, the changing perception and use of time and the coordination of everyday life (Rheingold, 2002; Ling, 2004; Castells, et al., 2006). By enabling self–organizing mobile social networks, mobile devices could widen the scope of existing issues and open new relevant questions. For instance, how education systems should address the need of citizens’ communication capabilities needed to exploiting the potential of ICT? And how governments and enterprises could exploit, through appropriate strategies, the power of self–organizing networks of active citizens? In other words, how to turn human networks into a resource for sustainability and not into a threat to political stability? Certainly, opportunities could turn into a nightmare for the humanity if the nature and properties of mobile social networks would be considered only from a technical and economic viewpoint, underestimating the importance of the human and social dimensions. End of article


About the author

Giuseppe Lugano received a Master’s degree in computer science from the University of Bologna (Italy) in 2003 and is currently a Ph.D. candidate in Cognitive Science and member of the Human Dimensions Research Group at the University of Jyväskylä (Finland). Since September 2004 he has been investigating mobile social networks and their impact on society as member of the Corporate R&D of TeliaSonera Finland. He is the author of the book Comunicazione mobile [Mobile communication], published in 2007 by the Italian publishing house Edizioni Cierre.
E–mail: gilugano [at] cc [dot] jyu [dot] fi



1. Lugano, 2007b, p. 2.

2. Wasserman and Faust, 1994, p. 20.

3. González, et al., 2008, p. 779.

4. Plant, 2001, p. 3.



Masaki Aida, Keisuke Ishibashi, Hiroyoshi Miwa, and Chisa Takano, 2004. “Structures of human relations and user–dynamics revealed by traffic data,” Lecture Notes in Computer Science, volume 3090 (February), pp. 492–502.

Réka Albert, Hawoong Jeong, and Albert–László Barabási, 1999. “The diameter of the World Wide Web,” Nature, volume 401, number 6749 (9 September), pp. 130–131.

John Arundel Barnes, 1954. “Class and committee in a Norwegian island parish,” Human Relations, volume 7, pp. 39–58.

Helen Bott, 1928. “Observation of play activities in a nursery school,” Genetic Psychology Monographs, volume 4, number 1 (July), pp. 44–88.

Julián Candia, Marta González, Pu Wang, Timothy Schoenharl, Greg Madey, and Albert–László Barabási, 2008. “Uncovering individual and collective human dynamics from mobile phone records,” Journal of Physics A: Mathematical and Theoretical, volume 41 (May), pp. 1–11.

Manuel Castells, Mireia Fernández–Ardèvol, Jack Linchuan Qiu, and Araba Sey, 2006. Mobile communication and society: A global perspective. Cambridge, Mass.: MIT Press.

Nathan Eagle and Alex Pentland, 2006. “Reality mining: Sensing complex social systems,” Personal and Ubiquitous Computing, volume 10, number 4 (March), pp. 255–268.

Marta González, Cesar Hidalgo, and Albert–László Barabási, 2008. “Understanding individual human mobility patterns,” Nature, volume 453, number (5 June), pp. 779–782.

Mark Granovetter, 1973. “The strength of weak ties,” volume 78, number 6 (May), American Journal of Sociology, pp. 1360–1380.

James E. Katz and Mark A. Aahkus (editors), 2002. Perpetual contact: Mobile communication, private talk, public performance. Cambridge: Cambridge University Press.

Elina Kuitto, 2001. “Lukiolaistyttöjen ja — poikien ystävyyskäytännöt sekä käsitykset ystävyydestä” [“The friendship practices and the conceptions of friendship among the high school girls and boys”]. Master’s thesis in sociology, University of Helsinki (Department of Sociology).

Paul F. Lazarsfeld and Robert K. Merton, 1954. “Friendship as social process: A substantive and methodological analysis,” In: Morroe Berger, Theodore Abel, and Charles H. Page (editors). Freedom and control in modern society. New York: Van Nostrand, pp. 18–66.

Lawrence Lessig, 2001. The future of ideas: The fate of the commons in a connected world. New York: Random House.

Nan Lin, 2001. Social capital: A theory of social structure and action. New York: Cambridge University Press.

Richard S. Ling, 2004. The mobile connection: The cell phone’s impact on society. San Francisco, Calif.: Morgan Kaufmann.

Markku Lonkila, 2004. “Phone notebooks as data on personal networks,” Connections, volume 26, number 1, pp. 53–61.

Charles Loomis, 1946. “Political and occupational cleavages in a Hanoverian village,” Sociometry, volume 9, number 4 (November), pp. 316–333.

Giuseppe Lugano, 2007a. “Designing for societal development: Enhancing Mobile Social Software through Social Capital,” iSoca 2007 Conference, Jyväskylä (Finland), October.

Giuseppe Lugano, 2007b. “Mobile social software: Definition, scope and applications,” eChallenges 2007 Conference, The Hague (Netherlands), pp. 1434–1441.

Giuseppe Lugano, 2006. “Understanding mobile relationships,” Human Centred Technology Workshop (HCT 2006; Pori, Finland).

Giuseppe Lugano and Pertti Saariluoma, 2007. “To Share or not to share: Supporting the user decision in Mobile Social Software applications,” Proceedings of the International User Modelling conference (UM 2007; Corfu, Greece, 25–29 July). Lecture Notes in Computer Science, volume 4511. Berlin: Springer. pp. 440–444.

Giuseppe Lugano, Jorma Kyppö, and Pertti Saariluoma, 2006. “Designing people’s interconnections in mobile social networks,” Proceedings of the First International Conference on Multidisciplinary Information Sciences & Technologies (InScit 2006; Merida, Spain, 25–27 October). Badajoz, Spain: Open Institute of Knowledge. pp. 500–504.

Peter V. Marsden, 2005. “Recent developments in network measurement,” In: Peter J. Carrington, John Scott, and Stanley Wasserman (editors). Models and methods in social network analysis. New York: Cambridge University Press, pp. 8–30.

Peter V. Marsden, 1990. “Network data and measurement,” Annual Review of Sociology, volume 16, pp. 435–463.

Stanley Milgram, 1967. “The small world problem,” Psychology Today, volume 2, number 1, pp. 60–67.

Andrew Newcomb and Catherine Bagwell, 1995. “Children’s friendship relations: A meta–analytic review,” Psychological Bulletin, volume 117, number 22, pp. 306–347.

Antti Oulasvirta, Mika Raento, and Sauli Tiitta, 2005. “ContextContacts: Re–designing Smartphone’s contact book to support mobile awareness and collaboration,” Proceedings of the Seventh International Conference on riendship and friendship quality in middle childhoodHuman Computer Interaction with Mobile Devices & Services, ACM International Conference Proceeding Series, volume 111, pp. 167–174.

Vilfredo Pareto, 1935. The mind and society. New York: Harcourt, Brace.

Jeffrey Parker and Steven Asher. 1993. “Friendship and friendship quality in middle childhood: Links with peer group acceptance and feelings of loneliness and social dissatisfaction,” Developmental Psychology, volume 29, number 4, pp. 611–621.

Malcolm R. Parks, 2007. Personal relationships and personal networks. Mahwah, N.J.: Erlbaum.

Sadie Plant, 2001. “On the mobile: The effect of mobile telephones in social and individual life,” at, accessed 17 April 2008.

Mika Raento and Antti Oulasvirta, 2005. “Privacy management for social awareness applications,” Proceedings of 1st Workshop on Context Awareness for Proactive Systems — CAPS 2005, Helsinki, Finland, pp. 105–114.

Mika Raento, Antti Oulasvirta, and Nathan Eagle, in press. “Smartphones: An emerging tool for social scientists,” Sociological Methods and Research.

Howard Rheingold, 2002. Smart mobs: The next social revolution. Cambridge, Mass.: Perseus.

Howard Rheingold, 1993. The virtual community: Homesteading on the electronic frontier. Reading, Mass.: Addison–Wesley.

Helen M. Richardson, 1940. “Community of values as a factor in friendships of college and adult women,” Journal of Social Psychology, volume 11, pp. 303–312.

Pertti Saariluoma, 2004. Käyttäjäsykologia [User psychology]: Ihmisen ja koneen vuorovaikutuksen uusi ajattelutapa. Porvoo: WSOY.

Linda Stone, 2005. “Continuous partial attention,” at, accessed 18 April 2008.

Marja–Liisa Viherä 1999. “People and the information society: The citizens’ communication skills and the opening of new prospects for the civil society,” Ph.D. dissertation, Turku School of Economics, Series A-1: 1999, pp. 337–354 (English summary).

Stanley Wasserman and Katherine Faust. 1994. Social network analysis: Methods and applications. Cambridge: Cambridge University Press.

Duncan Watts and Steven Strogatz, 1998. “Collective dynamics of ‘small–world’ networks,” Nature, volume 393, number 6684 (4 June), pp. 440–442.

Barry Wellman, 2001. “Computer networks as social networks,” Science, volume 293, number 5537 (14 September), pp. 2031–2034.

Barry Wellman, Peter J. Carrington, and Alan Hall, 1988 “Networks as personal communities,” In: Barry Wellman and S.D. Berkowitz (editors). Social structures: A network analysis. Cambridge: Cambridge University Press, pp.130–184.


Editorial history

Paper received 21 July 2008; accepted 10 October 2008.

Copyright © 2008, First Monday.

Copyright © 2008, Giuseppe Lugano.

Mobile social networking in theory and practice
by Giuseppe Lugano
First Monday, Volume 13, Number 11 - 3 November 2008

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2020. ISSN 1396-0466.