LOCATION BASED SOCIAL NETWORKING CHALLENGES AND SOLUTIONS AYESHA BEGUM MOUNIKA KOLLURI SRAVANI DHANEKULA.

LOCATION BASED SOCIAL NETWORKING

CHALLENGES AND SOLUTIONS

AYESHA BEGUM

MOUNIKA KOLLURI

SRAVANI DHANEKULA

OUTLINE

INTRODUCTION

EXAMPLES OF LBSN APPLICATIONS

PROS AND CONS OF LBSN

MOTIVATION

SELECTED PAPERS MEASURING USER ACTIVITY ON AN ONLINE LOCATION BASED

SOCIAL NETWORK.

PLACE RECOMMENDATION FROM CHECK-IN SPOTS ON LBSN.

LOCATION CHEATING ON LBSN.

COMPARISON & ANALYSIS

CONCLUSION

REFERENCES

INTRODUCTION:

LBSN is a location based service that utilizes location information to facilitate social networking.

LBSN is the convergence between location based services (LBS) and online social networking (OSN).

LBSN applications offer users the ability to look up the location of another “friend” remotely using a smart phone, desktop or other device, anytime and anywhere.

They allow users to check-in at places and share their location with friends, thereby providing a new facet of user online behavior.

EXAMPLES OF LBSN APPLICATIONS

PROS &CONS OF LBSN

PROS:It is the best means to share what we are doing and where we are at.Nearby, locations like restaurants, parks, zoo's can be easily found out.

CONS:LBSN lacks privacy of an individual and exposes user’s information.

MOTIVATION

A measurement study of user activity is needed in order to know how users connect with friends and also how they check- in at different places on online location-based social network with hundreds of thousands of users in it.

People generally don’t know the interesting locations when they go to new place. So, the social networking sites must be able to recommend places to them based on their area of interest gathered from previous check-ins.

Location-based social network services must be able to keep track of the users who cheat on their location information.

MEASURING USER ACTIVITY ON AN ONLINELOCATION-BASED SOCIAL NETWORK

The main aim of the paper is to present a measurement study of user activity on a popular Online Location Based Social Network.

The paper mainly investigates user activity by analyzing not only the number of friends user has, but also the number of check-ins made and the places visited.

EXISTING AND PROPOSED APPROACH

Existing Approach:User Activity measurement is focused mainly on number of friends

user has.

Proposed Approach:User activity is measured based on three factors

Adding Online Friends Making Check-ins Visiting new places

METHODOLOGY PROPOSED

Complete data set of users is collected through a public API.

User activity is measured, based on adding online friends, making check-ins and visiting new places. Probability distribution of these factors with respect to a user are plotted individually.

Based on the dates of both the earliest and the latest check-in that a user has made, account age and activity span of the user are estimated.

The user activity age and user account age are plotted graphically based on Complementary Cumulative Distribution Function (CCDF).

EXPERIMENTAL ANALYSIS

From the graphical analysis done in the paper, it states that

It appears easier and quicker to accumulate friends than to accumulate new places and check-ins.

It has been derived that an account which has been active for a longer period is more likely to accumulate more friends, check-ins and places than an account only active for a shorter amount of time.

User account life span decays faster than exponentially.

PLACE RECOMMENDATION FROM CHECK- IN SPOTS ON LBSN

The paper mainly discusses about a user-based collaborative filtering method to make a set of recommended places for a user, in which similarity of users is calculated and similar users’ records are used to predict places the user likes.

In addition to this, similarity of users check-in activities is calculated not only on their positions but on their semantics like shopping, eating, drinking, etc.

EXISTING APPROACH

In Collaborative Filtering Based Approach, an item is recommended to a user based on past information of the people with similar tastes and preference.

In Personalized Recommender Approach ,check-in information is crawled to generate a user/spot rating matrix. By predicting the interest of users in certain spots, this technique recommends places users have not been visited previously.

PROPOSED APPROACH

All of the existing methods do not consider the issue of the semantics of GPS location.

In the proposed method Firstly the names of the noted places are attached to GPS location

data and hierarchical category graph framework is built.

Recommendation of places is done by applying a typical CF approach that was not applied previously

DESCRIPTION ABOUT CHECK IN RECORDS

SINA microblog is used as a data source to collect user’s check in spots.

.

METHODOLOGY

Density-based clustering method cluster’s all of users’ check in spots into several regions.

In the next step, for each cluster, the gravity center of member’s position is calculated itto represent the position of the cluster .

Each cluster is annotated by using POIs database. Then a semantic hierarchical category-graph framework is applied to analyze users’ interests and similarity score between clusters.

Top-N similar users are selected and the users’ records are used for user-based collaborative filtering. Based on this, some unvisited places are recommended.

FIGURE SHOWING SEMANTIC HIERARCHICAL FRAMEWORK

EXPERIMENTS & ANALYSIS

SINA microblog, is used , among them, 268 users who checked in more than 25 times are selected.

In the clustering analysis ,the neighborhood radius ε actually equals to distance50 meters,

After the clustering, Foursquare database is used as POIs data to annotate clusters.

EXPERIMENTAL RESULTS WITH VARIOUS RECOMMENDER PLACES FROM DIFFERENT TOP-N SIMILAR USERS.

LOCATION CHEATING: A SECURITY CHALLENGE TO LOCATION-BASED SOCIAL NETWORK

SERVICES

The paper mainly discusses about the location-based mobile social network's which generally attract more no of user's, in order to provide real-world rewards to the user, when a user checks in at a certain venue or location.

This gives incentives for the cheaters to cheat on their locations.

Reasons for Location cheating

Lack of proper location verification mechanisms.

Loosely regulated anti cheating rules.

PROPOSED METHODOLOGY

Firstly, the threat of location cheating attacks is identified.Secondly, the root cause of the vulnerability is found out, and

possible defending mechanisms are outlined. Foursquare is used as an example to introduce a novel

location cheating attack. In addition to this, the foursquare website is crawled.

The crawled data is analyzed, in order to prove that the automated large scale cheating is possible.

CHEATER CODE

It is used to defend against the location cheating attacks. Function: It is to verify the location of a device by using the

GPS function of that device.

When a user claims that he/she is currently in a location far away from the location reported by the GPS of his/her phone, the check-in will be considered invalid and won’t yield any rewards.

CRITERIA FOR CHEATER CODE

Criteria used in determining location cheating in the cheater code is as follows Frequent check-ins: This rule prevents a user from checking in

frequently to get as many points as possible Super human speed: This rule limits location cheating by a single user

to a small geographic area. Rapid-fire check-ins: This rule stops a user from checking into multiple

venues in a small area and within a short time period.

DIFFERENT LEVELS OF CHEATING ATTACK:

Location Cheating Against G.P.S Verification: An attacker blocks the information provided by the GPS and feeds fake location information to the LBS application, thereby making the server believe that it is in fake location.

Crawling Data: By changing the ID in the URL, almost all of the user information and venue profiles is crawled. This is a serious security weakness and should be patched soon.

ILLUSTRATION OF LOCATION CHEATING

DIFFERENT LEVELS OF CHEATING ATTACK

Automated Cheating: Location coordinates of victim venues are found out by

computer program. List of venues that need to be checked-in are selected

automatically by analyzing the cheater code.

Cheating With Venue Profile Analysis Location cheaters gain intelligence from the venue analysis after

the crawling.

EXPERIMENT ANALYSIS OF LOCATION CHEATING ON FOUR SQUARE

Above Normal Level of Activity: High ratio of recent check-ins to total check-ins of a user indicates that it is likely a user plays tricks to stay in the recent visits list, which is a sign of cheating.

Below Normal level of Rewards: User having a large amount of check-ins but little rewards indicates that user is detected as a cheater.

Suspicious Check-in Patterns: Check-in pattern or history is examined to tell if a user is a location cheater through further analysis of the crawled data.

POSSIBLE SOLUTIONS BASED ON EXPERIMENTS

Location Verification Techniques: Address Mapping Venue Side Location VerificationMitigating Threat from Location Cheating Access control for Crawling Hiding information from profiles

COMPARISON AND ANALYSIS

Paper 1 Paper 2 Paper 3

Techniques focused on how the user activity can be effectively measured

Techniques discussed for extracting the check-in spots based on user’s interest

Techniques for enhancing the security of local information

Experiments performed onGowalla LBSN site

Experiments performed on SINA LBSN site

Experiments performed on Foursquare LBSN site

It highlighted the differences in the distribution of friends, check-ins and places

It recommends unvisited places by analyzing user’s interest

It provides better solutions to identify possible cheaters.

CONCLUSIONS

All the three papers discussed on location-based services which utilize the geographical position to enrich user experiences in a variety of contexts.

The papers conclude that Location based Features can effectively measure user activity, recommend unvisited places and also detect threat of location cheating attacks.

REFERENCES: Scellato, S.; Mascolo, C. Measuring user activity on an online location-basedsocial network Computer

Communications Workshops (INFOCOM WKSHPS), 2011 IEEE Conference on Topic(s): Communication, Networking & Broadcasting ;Components, Circuits, Devices & Systems ; Computing & Processing (Hardware/Software) ; Engineering Profession ;General Topics for Engineers (Math, Science & Engineering) ;Signal Processing & Analysis Digital Object Identifier: 10.1109/INFCOMW.2011.5928943 Publication Year: 2011 , Page(s): 918 - 923 Cited by 1

Hongbo, Chen; Zhiming, Chen; Arefin, Mohammad Shamsul; Morimoto, Yasuhiko Place Recommendation from Check-in Spots onLocation-Based Online Social Networks Networking and Computing (ICNC), 2012 Third International Conference on Topic(s): Communication, Networking & Broadcasting ;Components, Circuits, Devices & Systems ; Computing & Processing (Hardware/Software) Digital Object Identifier: 10.1109/ICNC.2012.29 Publication Year: 2012 , Page(s): 143 – 148

Wenbo He; Xue Liu; Mai Ren Location Cheating: A Security Challenge toLocation-Based Social Network Services Distributed Computing Systems (ICDCS), 2011 31st International Conference on Topic(s): Communication, Networking & Broadcasting ;Computing & Processing (Hardware/Software) Digital Object Identifier: 10.1109/ICDCS.2011.42 Publication Year: 2011 , Page(s): 740 - 749 Cited by 2

Thank You