Abstract
Mobile social networks have attained remarkable developments in the recent years. The rapid evolution of mobile technology provides a great opportunity for mobile social networks to share experiences, interests, opinions and personal information. Such platform can provide a variety of features for users to share data with others. Further, it can search the users with similar interests to establish and maintain communication. The user community of mobile social network started exploiting this platform. However, the success significantly depends on trusted members as well as service providers. Always there is a demand for sharing the user locations in the network of social activities which is specifically in hand held devices like smart phones. In fact, it is a major privacy concern. We first reviewed the existing trust model which is relevant to mobile social networks. Various location-based services were reviewed. Finally, the literatures are compared and areas for future work are identified.
Introduction
Mobile Social Networking (MSN) occurs in a virtual community, where people with similar interests converse and connect with one another through their mobile devices like Smartphone’s/tablets. MSN has evolved rapidly due to the popularity of Smartphone’s and internet all over the world. In the time frame between January 2016 and January 2017 [101], there was an increase of 30%, active users, which is about 581 million. The web social networks have been replaced by mobile apps, launched by various service providers. Due to tremendous success in the adoption of mobile devices, the MSN supporting apps have become an integral part of human lives. The traditional web-based social network sites, such as Facebook, Twitter etc., have created apps to give their real-time services and mobility. In addition, other MSNs such as Instagram, Foursquare and Strava allows end-users’ to build social communities through mobile features. MSN emerged as an ideal environment for establishing brand communities to share the information. MSN is a dynamic virtual network; as a result, there exists higher risk when mobile users try to interact with each other [1].
This paper mainly focuses on literature review of trust and services offered for user area oriented mobile networks with social context. The adoption of mobile social network arises several challenges. In general the fundamental issues exist for mobile social network are trust and repudiation. Identity management, location based service, incentive mechanism and human computer interaction. Mobile social network is an open platform which has major concern on trust and location based service compared to other services. Hence, various literatures relevant to this to the topic of interest are accounted for survey. The users of MSN do not have interaction and previous experience about other users, therefore evaluating the trust relationship becomes important between the mobile users in a social network. For example, in mobile commerce applications, the user needs to decide a service provider based on trustworthiness [2]. Further trust in MSN is more challenging than traditional social networking due to the mobility of users. In addition, it is dynamic and complex network infrastructure and also there is an absence of centralized environment. MSN users are in full swing, in exploiting this environment for delivering their services. The success of mobile services depends on the trust level being maintained mutually among the group as well as other network providers and the mobile nodes usually have constraints like limited energy, bandwidth and computational power [12].
Social network’s success rate increases when it offers services with trust as key element which acts as a metric to describe the actual status of the mobile node present in the network during transmission. Evaluation of interactions between mobile nodes is crucial for the development of trust, as it is widely accepted as a major component to identify as well as to validate the social relationships between the mobile nodes [13, 14, 15, 23]. A certain level of trust is critical for successful transmission and trust based framework is used to evaluate the delivery competency of the mobile nodes more accurately. Several trust protocols are also used for bias minimization and performance maximization in the network [16, 17].
The present mobile devices offer software level service in which information is provided through Location Based Service (LBS). This service uses information on a geographical position on their mobile devices. LBS are adopted in many applications such as the recommendation for social events in a city, e-promotions offered by retail shops and restaurants, in a location of people and mobile healthcare system. LBS in Smartphone’s should be enhanced in protecting users’ privacy. Location-based privacy is an important requirement when the user should have controlled decision regarding who can view their location. To overcome the privacy concerns with sharing of location, several researchers’ addressed and proposed different techniques. Hence there is a need to focus on trust and location-based services.
From the literature, it is noted that the social networks are first being utilized by individuals with their family as well as friends. The adoption of MSN exploited the potential of social networks to government and private sectors as a platform for their services. This open nature has made concern about trust and privacy for the individuals compared to other factors of MSN such as services, network model and interaction models, etc. Therefore, in this survey, service models are considered by matching the objectives of this work. This paper presents a comparative analysis of a various trust and location-based service models, properties existing in the literature, the key parameters and the strengths and limitations of these models.
The remaining sections of this paper are organized with Section 2 presents a detailed analysis of previous reviews related to the trust and location based network for social activities. In Section 3, an overview of various trust definition, trust types, trust properties, models, detailed review related to trust in MSN and its summary are addressed. Section 4 provides an overview of location-based services such as protocols and standards, classification, architecture, detailed review related to location-based services and its summary are addressed. The last section concludes the paper.
Overview
The trust in network is a concept majorly used in multi-disciplinary areas. It is not surprising that some reviews have been carried on trust. The Online social networking with mobile computing is a rapid succession to emerge as a new platform. The location-based services offer several advantages and also raise significant privacy concerns. In this paper, we attempt to summarise some of the important articles on trust and LBS.
In many engineering research, trust has been reviewed a considerable interest in the decision-making process. Significant efforts have been carried out earlier to address trust and trust metrics in various contexts. However, the issues concerned with how to model and quantify have not been addressed clearly. This paper [95] has been surveyed and outlined the concept of trust in the above-mentioned aspects. The survey mainly focuses on trust assessment, trust constructs, trust scales, trust properties, trust formulation and applications of trust. It further discusses how different trust can be mapped to different layers of a complex and composite networks, the applicability of trust metrics, models, etc.
Azer et al. [94] made a survey on trust for ad-hoc networks. In this survey, they attempted to focus on various goals, features and architecture for trust.
Mobile Adhoc Networks (MANET) [66] is a structure-less and dynamic network which consist of the mobile node without a physical link and does not have fixed topology schemes. A survey of different trust model schemes for MANET with their unique features, merits and demerits have been addressed [94, 97]. Suryanarayan and Taylor [96] focussed their review in P2P applications with a reputation based, a social network based and credential and policy-based trust. Golbeck [86] proposed a trust model in web-based social networks. In his work, he has presented two sets of an algorithm for trust inference, one for networks and the second one for continuous rating. Further, the computed trust inference has been applied in two applications:
Film trust is an application which combines trust models, social networks, movie ratings and reviews. In this approach, trust is used to personalize the website for display of recommendation for movie rating and also ordering the reviews based on its relevance. Trust mail is a client based email application which uses trust rating for each sender as a score for the message. In this approach, email user can sort the messages according to their trust rating.
Trust and reputation systems for internet application play a significant role in decision support. In their review, they derived a mechanism of trust and reputation for internet transactions to analyze the current trends and developments.
The localization techniques, which are offered in MSN, are used to share their locations and contents. This is referred as the location-based social network, which significantly associates physical and mobile devices. Bao et al. [100] attempted a systematic review and summarized their relations. They further presented a comprehensive survey in analyzing the data source used, the methodology employed to generate a recommendation and objective of the recommendation.
Krumm [98] attempted to survey the computational location privacy, meaning computation-based privacy mechanisms that treat location data as geometric information. The privacy-preserving algorithms such as anonymity and obfuscation as well as privacy-breaking algorithms have been addressed the geometric nature of the data. The paper reviews about peoples’ attitudes, location privacy, and computational threats on leaked location data. Further computational countermeasures for mitigating these threats are also mentioned.
Raper et al. [99] reviewed a selected set of location-based services (LBS) that have been published in the various research literatures. This paper mainly focuses on mobile guides, transport support, gaming, assistive technology and health. The review illustrates the enormous diversity of forms in which LBS are appearing and the wide range of application sectors that are represented.
This paper is organized based on two criteria: trust and location-based service in MSN as shown in Fig. 1. Trust is categorized to four sources (i) Trust Origin, (ii) Trust Type, (iii) Trust Properties, and (iv) Trust Models. Location-based services are categorized to (i) Protocols and Standards, (ii) Classifications, (iii) Characteristics and (iv) Architecture.
Taxonomy of trust and location-based service in a mobile social network.
An overview of trust.
Trust is a multidisciplinary concept being used in applications such as sales and marketing, philosophy, psychology, economics, sociology and social identity and so on [85].
In general, trust can be defined as confidence relation between trustor and trustee. Trust definition has been adopted in a different application with different viewpoints such as reliability, availability, reputation, risk, confidence and quality of service. The overview of trust is mentioned in Fig. 2.
Trust origin
In sales and marketing [7, 73], trust is mainly focused on belief, willingness to believe and confidence in nature of construct. A trustee is defined by its characteristics such as honesty, competency benevolence, etc.
Trust in sociology is a recurrent subject of discourse in the various literature [24, 56]. Trust has become the focus of great attention to work guided by experience in efforts to identify its cause and effect in social life.
In psychology [5, 33], Trust has been considered as different stages of psychological development. For example: At infants’ stage, trust or distrust has experienced to an environment in which they live. If they live reasonable and dependable, trust is acquired. On the other hand, if they live with abuse and dangerous environment, distrust is developed.
Trust in economics is treated as how human exhibit his behaviour in reducing the cost of the transaction. The optimum level of trust may provide a great opportunity for efficient market. In contrast, the economic opportunity may be lost if there is a lower level of trust. In certain cases, a higher level of trust may sometimes provide opportunities for vulnerability.
Trust in a computing system is considered as to how computer systems can consistently behave expectedly. It has completely relied upon a set of specific policies. If the policy exists, it is considered to be trusted or else it is considered as distrusted.
Trust types
Different types of trust can be considered from the viewpoint of various disciplines.
One of the trust types is the calculative trust type in which the trust is the result of calculation to increase the participant’s stake in the interaction.
The continuous bonding between trustor and trustee is another type which is called as relational trust. The information come out from such bonding is the fundamental factor for relational trust.
The safety and ease are other factors which decide the emotional content of trust which depends on a trustee. In psychology, the direct relationships among persons produce the emotional aspect as the outcome.
The trust which is the outcome of rational behaviour and reason is of cognitive type. Three forms of factors of social capital affect cognitive trusts are suggested according to social capital theory: 1. Trustee’s obligations to the trustor, 2. Norms and sanctions and 3. Information channels [84]. Specifically, cognitive trust is increased by positive referrals within the relations of social networks [26, 27, 29, 30, 58, 59, 60, 61].
The environment at the institution, smooth relations among members and punishments for misbehaviour describe the characteristics of the institutional trust.
The institutional trust reveals that, during the lifetime, opinion of the people and expectations about trust on other people differs widely. It is sometime discussed as dispositional trust which is commonly applied to every individual of the institution without any barrier and conditions [35, 62, 63].
Trust properties
Trust is context specific in its scope. It is related to psychological and social in nature. It’s level may rise or fall according to the experiences. It disappears over time. Old experiences diminish over time and become obsolete in nature. Hence new incidents are dominant than old one and prioritized as important in deciding the trust characteristics. This dynamic characteristic is very much helpful in designing models in computer engineering.
In a propagative trust, person 1 trusts person 2 and person 2 trusts person 3. Person 1 does not know person 3. However person 1 can trust person 3 based on how much amount person 1 trust second person and in turn how much second person believes third person. Hence this kind of trust is transitive.
If first person trusts second one and second person trusts third, it does not match that there is a trust by first one with third person. In this case, trust is not transitive. Though there is propagation relationship with transitive, the reverse is incorrect. Here there is a misunderstanding about transitive and propagative trusts. The trusts and distrusts of propagative in nature can be made with an individual even though the person is unknown. This is possible in social chain. Because a trust made by a second person with third first person helps the first person to have a trust on third person through the trust made by 2nd person with 3rd person.
Trust differs in the social chain as it is subjective. Person 2 gives an opinion on a film. If the first person has a good thought on person 2, he may believe the review of person 2. But person 3 may differs the opinion of second person and may not consider the review of second one.
A member may believe another more than the reverse trust by the later one. It is called asymmetry in trust. The level of trust increases when they have more mutual interactions. It is a personalized relationship in asymmetry. Opinion, views, believes and confidence levels are the reasons for asymmetry in trust.
Trust is self-reinforcing. Members act positively with other members whom they trust. Similarly, if the trust between two members is below some threshold, it is more highly unlikely that they will interact with each other, leading to even lesser trust on each other.
Trust takes a long time to build, but a single high-impact event may destroy it completely. This aspect of trust has received even less attention in computer science [46, 47, 48, 49, 50, 51, 52, 53, 54].
Trust models
There are different methods of trusts utilized [85] in a social mobile networks which can be broadly classified into statistical and machine learning techniques, heuristic-based techniques, and behavior-based techniques. Statistical and machine learning techniques provide a strong mathematical concept for the managing of trust. Heuristics-based techniques focus on practical models for implementing trusted system. Behavior-based techniques mainly focus on how user behaves in their group [25, 55, 56, 57]. Some of the important techniques which are statistical methods are Bayesian systems [103, 104] and belief model [105, 106, 107]. In Bayesian technique, the binary rating for dishonest or honest are helped to judge trust by statistically enhancing density functions with beta probability. The trust has been computed based on the user’s belief with related to the truth of evaluation in a belief model. The computation and prediction of trust based on machine learning technique make use of techniques such as Artificial Neural Networks and Hidden Markov models.
As the statistical and machine learning techniques are highly complex, researchers have moved towards heuristic-based techniques. The solutions [108, 109] target to describe a robust, real time, simple to construct and understand the management approach for trust. The author [110, 111] presents a behavior-based technique, in which trust is judged with the help of the communication behavior of users. Behavior trust is calculated using conversation trust and propagation trust. Conversation trust identifies the time and frequently two users communicate with one another. Longer and frequent conversation indicates more trust between each other. Propagation trust refers to the propagation of information from one user to various other users indicating the high degree of trust in the information.
Literature review
Figure 3 shows the taxonomy of trust in MSN based on various characteristics. These characteristics are generalized and may vary in different ways in different environment. These characteristics has been adopted and discussed in different literature are mentioned below.
Taxonomy of trust.
Domingo-Ferrer et al. [67] proposed a new protocol which enables a private relationship in a social network. This approach allows accessing the resource using indirect relationships without a trusted third party as a mediator. This approach has been implemented using homomorphic encryption which prevents resource owner. This scheme restricts to learn the relationships and trust levels between the social network users to collaborate in accessing various resources. This approach, provides privacy concern of the users who refuse to collaborate have been greatly minimized. This approach can be highly scalable and deployable in social networks.
Hao et al. [68] have proposed a new fuzzy inference mechanism, namely MobiFuzzy Trust. MSN allows a mobile user to find users of similar interests to establish communication and share the information. MSN is public virtual space where the available information is not trustworthy. The authors have proposed a new mechanism mainly MobiFuzzy Trust for inferring the trust between the mobile users in MSN. In this paper, they constructed an intersection of the prestige of users’ location, time and social context. Further, they developed a trust model to evaluate the trust value between mobile users. Finally, they adopted the fuzzy linguistic technique to represent the trust between mobile users and further they enhanced the trust. Through their experimentation on real-world mobile data set, they demonstrated that MobiFuzzy Trust could efficiently infer the trust.
Liang et al. [69] proposed a Trustworthy Service Evaluation (TSE) system for service-oriented mobile social networks (S-MSNs). This approach allows one or more users to communicate with a service provider who offers various services. In this context, maintaining a trust relationship between service provider and the user is very much important. This approach allows the service provider to maintain TSE which collects and maintain reviews about its services without any trusted third party. The service reviews given by various users can greatly help the interested user to make a wise decision for selection of the service. They further identified three possible review attacks on various services namely linkability, rejection and modification attack. To overcome the above-mentioned issue they designed a secured TSE in S-MSNs.
Wang et al. [72] presented a trust outcome for enhancing the trust in network for social activities based recommendation system on network companies. In a service-oriented platform, it has a huge amount of operators who offer more number of facilities to users. In a traditional recommendation system, less amount of data usage and trust improvements are the major issues. Recently, to overcome this issue, recommendation system adopted online social network to collect information from the social network users for generating the final recommendation. However, this approach lacks to access the trust between participants’ without any direct interaction. In this paper, these issues have been overcome by taking participants’ personal characteristics and mutual relationship which includes belief, exchange of social information among two users. In addition, they proposed a new probabilistic approach known as a social trust to infer social context in a social network. The algorithm proposed by them will also take information updates in a social network into account for trust evaluation. Their experimental result demonstrates that their method is efficient as well as dominant with respect to various trust outcomes in social network based recommendation system.
The mobile service provider is linked with two types of user groups. One is within network arrangement and another one is the users. Mobile services are insecure in these types users. To eliminate the problems, Scudellari et al. have suggested a trust-based model for retaining privacy called SMILE: trust for social activities for mobile which is encounter based approach [90]. In this method, a trust is established based on service encounters. It is a centralized and secured approach. The added features of SMILE model are given below:
Passively exchanging the cryptographic key with nearby peers. Periodically updating hash key to central and coordinating services. A user sends a secured key and label along with a message to the server with corresponding key pairs. A server is designed to forward encrypted message to all users using a key hash. Only encounter participants are allowed to decrypt the message.
Dhami et al. [91] created a design to know the outcome of security, trust and privacy concerns for exchanging the willingness based information in social networking. In their model they took 265 Facebook members in which 246 were used for a study. In those samples, the user with information control and profile security are called trusted group of users. In that survey, they decided that the social network over internet is secure and there is a definite relationship between privacy and trust.
MSN is one of a random and as well as risky environment. Hence calculate or express the trust at user’s point of view is a major concern. In this point of view, among mobile users, building valid and mutual understanding in the case of trust is a major concern.
Updates and advancements with respect to technical aspect, mobile gadgets are named as Smartphone’s. Smartphone’s offer good platform for mobile social network and communication for information sharing. As discussed a short while, the information through MSN is insecure and not trust worthy. However the trust is an important element in mobile network with social content without any doubt. Behaviour graph for social activities as expressed by Chen et al. [93] deals the implicit, i.e., ego-i graph along with a procedure (algorithm) for implementing ego –i graph. In that, the group based trust values are calculated using according to various contact persons, attributes and relationships among such contacts. By accumulating mutual group based trust weightage, the authors find a solution to form a group of trusted users as clusters. The evaluation is done at last to exhibit the characteristics of trust model for cluster group which is done for MSN network.
Qureshi et al. proposed a framework for building trust-based community in Peer to Peer (P2P) mobile social network [87]. P2P mobile networks greatly attracted the internet in the recent times. Users in P2P mobile social networks associate with each other and exchange the information without depending on centralized infrastructure. In this approach the users synchronize with their internet can initiate group or communities and share various resources. The membership in a group or community the MSN will be provided based on recommendation provided by the existing member of that group without any centralized authority. To overcome this, they proposed a decentralized P2P framework for exchanging the trusted information based on dynamicity aware graph relabeling system (DAGRS). This approach adopts lightweight trust model to identify trustworthiness of the user. Further, this model intended to create a community of users and reduced interaction with an untrusted user. To prove the effectiveness of proposed framework, they implemented in three different environments: campus network, shopping mall networks and city street networks.
The rapid development of mobile technology, the social networking services have been integrated with mobile phones. This provides a great opportunity for mobile social networking application to be more convenient and overcome the limitation by the environment. Social interaction provides a great opportunity to solve many problems directly or indirectly through friends. MSN is such kind of platform which integrates mobile technology and social networks. Social networking is a large and complex network. Diijkstras’ algorithm is a classical algorithm for finding the shortest path between two nodes. In case of complex social networks, the complexity is enormous. To improve this situation, Wang et al. has proposed a novel algorithm which decides an efficient routing by three-dimensional factors whereas Diijkstras’ algorithm involves only one factor for an edge in the graph [88]. In this paper the major focus to find the suitable target by matching the trust path. The worst case is if all the nodes and edges are traversed Diijkstras’ algorithm results with the time complexity of
The developments of wireless technologies like wifi, 3G, 4G, etc., in mobile devices provides great opportunity to exploit various internet services. The users greatly utilized these services to share messages and files posted by others irrespective of location. The messages which are exchanged may not be accurate. Lu et al. [89] developed a trust-based authentication system for MSN. To validate the creditability of messages exchanged among the users have been validated by the rank issued by his/her friends through social networks. In this approach, they constructed a trust graph in Facebook and established the credibility of mobile users by applying transitivity on a single path computed on trust graph. They further exploited the importance of multipath to overcome more suitable attacks which are not addressed by single path. The approach to artificial social trust network has experimented and then validated on real social network data.
This section provides a detailed comparison and performance analysis of trust in mobile social networks in terms of techniques used with its strengths and limitations as shown in Table 1. Most of trust models use relational aspect of trust based on direct past interaction. Use of the other trust characteristics like computative, cognitive and dispositional has been limited. The mainstream of the trust models are behavior based. Behavior trust is calculated based on conversation trust and propagation trust. Table 1 show that different attributes of trust get very light important as discussed in survey, like asymmetric, individual characteristics, as well as variations related to occurrence.
Comparison summary of trust in a mobile social network
Comparison summary of trust in a mobile social network
For more than a decade the mobile technology provides the facility for the people to use a cell phone to get directions, used to track their friends, to focus on special objects, to find nearest places like restaurant, hotels, hospitals, etc. The present generation Smartphone’s include the functionality of Global positioning system and other products as there is a tremendous growth in the Smartphone users with internet facilities [102]. The traditional online social network notifies others about who the user is. The location-based service in mobile can inform others about where a user is. This application greatly helps to associate technology and physical existence of the people.
During 2001, mobile service providers in US and Japan tested the features of location tracking and gained successful enhancements. Alongwith features of satellite communication such as GPS, GALKEO other arrangements have provided greater benefits to the location-based system. Further location details are retrieved from operating frequency of services, sensor values, Bluetooth, Wimax and Wireless local area networks.
Location-based protocols and standards
Many organizations played a significant role in developing the standard for location-based services like OMA (Open Mobile Alliance) etc. Such companies have offered several protocols and standards which are as mentioned below: Mobile Location Protocol describes the prime functional specifications and it is an a protocol for applications to obtain positional information of mobile station. It acts as an interface between location server and client. To overcome Location Interoperable Forums (LIF) mobile location protocol, OGC has introduced an OpenLS service that addresses the geospatial interoperability problems. Internet Engineering Task Force (IETF) has developed the standard Geopriv for sending the positional information on mobile data.
Classification of location-based services
Based on the positional information of mobile devices, Location-based supports are categorized as mentioned below.
An urget service is used to transfer the exact location information of mobile user automatically for emergency assistance. Navigation services are used to get the direction in their current geographical location. This service is used to determine the position based on prediction and routing calculations to locate friends, family members, workers, etc. for improved communication.
In Information/Directory services, the location, time, specificity and user behavior will be distributed based on the content of mobile terminal devices.
The following are the categories of information/directory services:
Travel service: It provides the information to the tourist about the city, notification of the place of interest, transportation service, etc. Mobile yellow pages: This service is provided to a mobile user based on the request to get information about nearby facilities. Infotainment service: These types of services are used to get user area oriented contents belong to multimedia type and community based events. Advertisements refer to publicity and business activities. Short Messaging Services (SMS), Mobile banners and Proximity triggered advertisements are the different forms of advertising services.
The different location-based services are classified as: individual-oriented, system dependent, direct as well as indirect, availability of user details, contents related to static and dynamic information and lastly about location accuracy [102]. In person-oriented location-based services, typically how the user location detail is achieved and utilized. The device-oriented location-based services are external to user and are not controlling the service.
The user details requested by user on the go or available readily to location-based services. The scenarios of mobile network available by means of user activities and network. Level and type of interactions depend on mobility scenario. Static information sources are collected from historical buildings and landmarks, places of attraction, hotels and restaurants and maps whereas dynamic information sources from weather, traffic and road conditions that changes with time. The source of details of location is offered by the network customer or the resource. The correctness of the mobile locations can be ensured with the help of technology being used.
Architecture of location-based services
Location-based service architecture mentioned below consists of following entities:
The architecture of location-based service is mentioned in Fig. 4.
Architecture of location-based services.
Taxonomy of location-based service.
The taxonomy of location-based service has been organised by considering three criteria namely protocols and standards, categories and types. Figure 5 shows the taxonomy of location-based service. The above criteria adopted in various literatures have been discussed below.
Li et al. [74] developed a system which tracks user location automatically as well as tested it on leading location-based social networks including Wechat, Skout, and Momo. The results are checked for its capability and competency through a three-week real time testing on 25 interested users and shown that it can geolocate any target with correctness and ability to identify top 5 user areas. Lastly, they also developed a platform which explored a reference system and location types to reduce the attack. The outcome helps to support the LBSNs with more number of users as a reference for security.
Wei et al. [75] presented a method which offers adjustable and security included mOSNs model with privacy preserving characteristics. Many types of user area oriented applications were helped by MobiShare, wherein it offers sharing of location for un-trusted strange user as well as trusted. It offers user assistance and controllability as desired by users. The network server for social contents and server of the user location do not have an idea of identities of users. At the time of suspicious attacks, the privacy of user location is maintained highly.
Cherng and Aritsugi [76] proposed a method which offers no leak to systems which shares location and protects privacy. The commonly available geographical information is transformed as region maps of individual details. The method also offers no connection among actually available geographical data and location of users to protect suspicious attacks.
Shankar et al. [77] presented Social Telescope, a user location oriented support which arrange indexes and prioritize the user locations based on communication with respect to position. The user locations are trained using data from twitter and the resulting system is turned into a search engine with location as key input. Methods like page based ranking by Google Search, expert oriented search by Zagat and customer ratings based search by Yelp are taken for comparison by authors against their suggested model for evaluation. With very lower costs, the user and location oriented network service offer results which are as relevant to the existing methods.
Chow and Mokbel [78] focused on the ubiquity of mobile devices with global positioning functionality (e.g., GPS and AGPS) and Internet connectivity (e.g., 3G and Wi-Fi). There are two methods of location based services available namely the continuous LBS and snapshot LBS. In continuous LBS, the network owner requires location details of the user either at regular interval or on need basis. Whereas in snapshot LBS, the network user requires to inform the current location to network provider about the required area of use. Method to Protect privacy of user location is easier for snapshot LBS than the constant LBS approach as the later one is a tough one. Because the snapshot method use the spatial and temporal relationships, it has higher possibilities of occurrence.
A literature review which deals the correlation between technological data, identity and location information are done by Schwartz and Halegoua [79]. Next, they identified and characterized the special characteristics by taking data from popular social media network giants like Facebook, Instagram, and Foursquare. The solution is finally made by them by means of analyzing the geographical based coding information.
Zhao et al. [80] created a method to identify the factors which are responsible to investigate location-based social network users aim to reveal the information related to location in China. Besides, justice theory is used to identify the influence of privacy interference methods utilized to enhance the user awareness in LBSN websites, including option for incentives, promotion by means of inter relation, privacy policy and control. The predefined design based model evaluation is done by connection based location identification and reveal the data on them.
Zhu and Cao [81] proposed a method called APPLAUS which is a Bluetooth associated cellular model which cooperatively create location proofs as well as it updates its own server. Cellular gadgets use the pseudonyms which are changing at regular interval to defend the source location safety from one another as well as from the un-trusted server. The authors also developed user-centric location privacy model in which individual users evaluate their location privacy levels in real-time and decide whether and when to accept a location proof exchange request based on their location privacy levels. APPLAUS can be implemented with the existing network infrastructure and the current mobile devices and can be quickly deployed in Bluetooth enabled mobile devices with little computation or power cost.
Noulas et al. [82] examined how venue discovery behavior characterizes the large check-in datasets from two different location-based social services like Foursquare and Gowalla by using large-scale datasets containing both user check-ins and social ties. The investigation shows that, across 11 cities, between 60% and 80% of users’ visit are in locations which were not attended in recent one month. They then show that, by applying assumptions related to user location, familiar and standard filtering algorithms, including latent space models, do not create best class model suggestions. Finally, a new approach was anticipated with respect to the personalized dynamic user location graph by flawlessly mixing social network and location visit information which is arrived between 5% and 18% updation over rest of the methods.
Anisetti et al. [83] have presented an approach which offers geographical location as well as mobility prediction both at network and service level which does not require any variation to the available mobile network resources, and is completely done on the mobile network side, enabling as stable design than other positioning systems with respect to location spoofing and other terminal-based security attacks. The method was related to a novel database correlation approach over received signal strength indication (RSSI) data and provides a geolocation and tracking technique based on advanced map and mobility based filtering. The performance of the geolocation algorithm has been carefully validated by extensive experimentation, carried out on real data collected from the mobile network antennas of a complex urban environment.
Summary
Table 2 shows the comparative analysis of the methods used for location-based security in mobile social networks. New techniques and systems were developed to provide security in the network by tracking the locations such as location-based search engine, privacy-preserving location sharing system and a privacy-preserving location proof updating system, etc. Using these techniques, several security issues in the network can be solved. From Table 2 it is observed that to provide security in social networks, each of the above-specified methods can able to detect the malicious behaviour of network based on location accurately and efficiently.
Comparative analysis of location-based services in mobile social networks
Comparative analysis of location-based services in mobile social networks
This paper provides an overview of a research article to trust and location-based services in mobile social networking. It is obvious that there exist some merits and limitations in mobile social networks. The issues like trust recommendation system and location privacy preservation would serve as a reference and support for understanding the current trends in research perspectives. Further, the summary of various techniques is compared concerning their model, performance, strengths and limitations. From our analysis, it is observed that each of the solutions has its advantages and limits. This signifies a research space in the existing solutions that can provide enormous opportunities to investigate in many aspects.
Footnotes
Authors’ Bios