Combating Friend Spam Using Social Rejections

Transcription

1 Combating Friend Spam Using Social Rejections Qiang Cao Duke University Michael Sirivianos Xiaowei Yang Kamesh Munagala Cyprus Univ. of Technology Duke University Duke University

2 Friend Spam in online social networks (OSNs) 2

3 Friend Spam in online social networks (OSNs) Friend spam: unwanted friend requests Fake account 3

4 Friend Spam in online social networks (OSNs) Friend spam: unwanted friend requests Ø Degrade user experience (e.g., annoying) Ø Introduce false OSN links Fake account 4

5 False OSN links are harmful Pollute the underlying social graph Ø Detrimental to social search and online ad targeting Ø Jeopardize online privacy and safety 5

6 False OSN links undermine the effectiveness of Sybil defense The defense relies on genuine social links Ø SybilLimit [S&P 08], SybilRank [NSDI 12] Ø # undetected Sybils (fake accounts) is bounded to O(log V ) per link between Sybils and legitimate users OSN links Non-Sybil region Sybil region 6

7 Existing counter- measures Privacy sexings for OSN users Ø Restrict requests only from friends of friends Ø Subtract from the openness of the OSN 7

8 Existing counter- measures Privacy sexings for OSN users Ø Restrict requests only from friends of friends Ø Subtract from the openness of the OSN Spam request filtering using machine learning (ML) Ø Facebook Immune System (SNS 11) Ø Individual user features are manipublable 8

9 Rejecto: Combating friend spam using social rejections 9

10 Observation: the cost of connecting to real users False OSN links come with social rejections Legitimate users Friend spammers 10

11 Observation: the cost of connecting to real users False OSN links come with social rejections Legitimate users Friend spammers 11

12 Observation: the cost of connecting to real users False OSN links come with social rejections Ø Social rejections: rejected, ignored, and reported requests Ø Spam requests are less likely to be accepted Legitimate users Friend spammers 12

13 Observation: the cost of connecting to real users False OSN links come with social rejections Ø Social rejections: rejected, ignored, and reported requests Ø Spam requests are less likely to be accepted Many rejections Legitimate users Friend spammers 13

14 Live fake accounts in the wild Each has a significant number of pending requests Ø Fake Facebook accounts from underground market Ø More measurement results in the paper Number of requests Pending requests Friends Anonymized fake account ID 14

15 How reliable is social rejection? AXackers inevitably trigger rejections Ø Disproportionally large number of accounts and requests Ø Requests inevitably hit cautious users Rejection towards innocent users is non- manipulable Ø A rejection is guarded by a feedback loop between the request sender and the receiver Ø Legitimate users rarely receive rejections Ø Fundamentally different from negative ratings on online services (e.g., YouTube) 15

16 Challenges to use social rejection AXack strategies Ø Collusion: fake accounts collude to accept requests Ø Arbitrarily boost the request acceptance rate of an individual account Ø Self- rejection: mimic legitimate users rejecting others Ø Whitewash the part of rejecting fake accounts System challenge Ø Gigantic user base with enormous requests and rejections 16

17 Rejecto in a nutshell A strategy- proof formulation Ø Graph cut on a rejection- augmented social graph Ø Low aggregate acceptance rate of the requests from spammers to legitimate users An effective and near- linear algorithm Ø Based on the Kernighan- Lin (KL) algorithm [The Bell System Technical Journal, 1970] A scalable implementation Ø Layered on top of Apache Spark [Zaharia et al. NSDI 12] 17

18 Outline Key insight System design Evaluation 18

19 Rejecto s formulation of spammer detection Main idea: put spamming accounts into groups 19

20 Rejecto s formulation of spammer detection Main idea: put spamming accounts into groups F ( H, S) Aggregate acceptance rate (AAR) F ( H, S) +!" R ( H, S) H S 20

21 Rejecto s formulation of spammer detection Main idea: put spamming accounts into groups F ( H, S) Aggregate acceptance rate (AAR) F ( H, S) +!" R ( H, S) H S 21

22 Rejecto s formulation of spammer detection Main idea: put spamming accounts into groups F ( H, S) Aggregate acceptance rate (AAR) F ( H, S) +!" R ( H, S) Fake accounts cannot arbitrarily improve AAR H S 22

23 Rejecto s formulation of spammer detection Main idea: put spamming accounts into groups F ( H, S) Aggregate acceptance rate (AAR) F ( H, S) +!" R ( H, S) Fake accounts cannot arbitrarily improve AAR H S 23

24 Spam requests lead to a low aggregate acceptance rate Lower than the requests from a set of legitimate users Ø Spam requests are less likely to be accepted 24

25 Spam requests lead to a low aggregate acceptance rate Lower than the requests from a set of legitimate users Ø Spam requests are less likely to be accepted 25

26 Spam requests lead to a low aggregate acceptance rate Lower than the requests from a set of legitimate users Ø Spam requests are less likely to be accepted 26

27 Spam requests lead to a low aggregate acceptance rate Lower than the requests from a set of legitimate users Ø Spam requests are less likely to be accepted 27

28 Spam requests lead to a low aggregate acceptance rate Lower than the requests from a set of legitimate users Ø Spam requests are less likely to be accepted A small AAR ratio cut 28

29 A graph cut model Augments a social graph with rejections Ø Directed rejection edges Finds the cut with the minimum aggregate acceptance rate (MAAR) Ø Graph partitioning based on requests and rejections Iteratively cuts off groups of suspicious accounts Ø Prunes their links and rejections from the social graph 29

30 A graph cut model Augments a social graph with rejections Immune to collusion and self-rejection strategies Ø Directed rejection edges Finds the cut with the minimum aggregate acceptance rate (MAAR) Ø Graph partitioning based on requests and rejections Iteratively cuts off groups of suspicious accounts Ø Prunes their links and rejections from the social graph 30

31 Outline Key insight System design Evaluation 31

32 Finding the MAAR cut MAAR cut is NP- hard is challenging Ø Reduced from MIN- RATIO- CUT problem [Leighton & Rao, JACM 79] Ø Detailed reduction in the paper Existing work on cut- based problems in undirected graphs Ø State of the art: O(log V ) approximation algorithms with complexity of O( V 2 ) [Madry, FOCS 10] 32

33 Finding the MAAR cut MAAR cut is NP- hard is challenging Ø Reduced from MIN- RATIO- CUT problem [Leighton & Rao, Ø The JACM 79] approximation factor O(log V ) is too loose Ø Detailed Ø O( V reduction 2 ) complexity in the is paper prohibitive Ø Do not support parallel graph processing Existing work on cut- based problems in undirected graphs Ø State of the art: O(log V ) approximation algorithms with complexity of O( V 2 ) [Madry, FOCS 10] 33

34 Our approach: an effective and efficient search algorithm Finds a MAAR cut by interchanging misplaced nodes Ø Based on the Kernighan- Lin (KL) algorithm Ø O( V ) complexity Ø Can scale up to multimillion- node social graphs 34

35 Our approach: an effective and efficient search algorithm Finds a MAAR cut by interchanging misplaced nodes Ø Based on the Kernighan- Lin (KL) algorithm Ø O( V ) complexity Ø Can scale up to multimillion- node social graphs 35

36 A primer on the Kernighan- Lin (KL) algorithm Searches a balanced cut in undirected graphs Ø Minimizes #cross- partition edges Ø Reduces cross- partition edges by swapping nodes Ø Fudiccia et al. improved to O( V ) [DAC 82] Ø Widely used in VLSI layout design U V-U 36

37 A primer on the Kernighan- Lin (KL) algorithm Searches a balanced cut in undirected graphs Ø Minimizes #cross- partition edges Ø Reduces cross- partition edges by swapping nodes Ø Fudiccia et al. improved to O( V ) [DAC 82] How to use KL to find the MAAR cut? Ø Widely used in VLSI layout design Ø Additional directed rejection edges Ø Non- linear MAAR objective function U V-U 37

38 Transforming the MAAR cut problem Convert to a set of bipartition problems Ø Each with a parameterized linear objective function Ø Rejection and social links can be unified F ( V S, S) ( ) +!" R ( V S, S) F V S, S F ( V S, S) k!" R ( V S, S) Solvable by KL after unifying the rejections and OSN links according to the parameter k 38

39 Why can we do the transformation? The MAAR cut is an optimal solution to one of the converted family of bipartition problems Ø The converted problem is determined by the MAAR cut ratio k* Theorem: In a rejection- augmented social graph, if the cut C * = U *,U * is the minimum aggregate acceptance rate (MAAR) cut, and F ( U *,U * ) = k * (k* > 0),!" * R U,U * C* is the optimal solution to the bipartition problem that minimizes. ( ) k * R!" U,U F U,U 39

40 Optimization and implementation Support seed pre- placement to reduce false positives Ø Seeds of both legitimate users and spamming accounts Prototype on Apache Spark Ø Distribute the large social graph to workers Ø Keep only a tractable set of algorithm states on the master 40

41 Outline Key insight System design Evaluation 41

42 Evaluation Extensive simulations on real social networks Ø Sensitivity analysis Ø Resilience to axack strategies Ø Compared to VoteTrust Simulations under Sybil axack Ø In- depth defense with social- graph- based Sybil defense A Rejecto prototype on an Amazon EC2 cluster Ø Performance analysis on large graph processing 42

43 Rejecto is insensitive to spam request volume Request flooding axacks on a Facebook sample graph Ø Fake accounts connect with each other as normal users do Precision/recall Rejecto VoteTrust Number of requests per fake account All fake accounts send out spam requests Precision/recall Rejecto VoteTrust Number of requests per fake account Only half of the fake accounts send out spam requests 43

44 Precision/recall Rejecto is insensitive to spam request volume Request flooding axacks on a Facebook sample graph Ø Fake accounts connect with each other as normal users do Rejecto uncovers fakes behind 1 1 the actively spamming ones Rejecto Rejecto VoteTrust VoteTrust Number of requests per fake account Number of requests per fake account All fake accounts send out spam requests Precision/recall Only half of the fake accounts send out spam requests 44

45 Rejecto is resilient to a]ack strategies Our MAAR cut model is immune to manipulation Precision/recall Collusion strategy to form dense connections among fake accounts 0.6 Rejecto 0.4 VoteTrust # of non-attack edges per fake account Precision/recall Self-rejection strategy to let half of the fakes reject the rest as legitimate users do Rejecto VoteTrust Self-rejection rate among fake accounts 45

46 Rejecto and social- graph- based Sybil detection form a defense in depth Rejecto makes fakes hard to get additional links Ø Defense in depth with SybilRank Area under the ROC curve Facebook ca-astroph Number of accounts removed by Rejecto Improvement 46

47 Rejecto can handle multimillion- user social graphs Performance on an EC2 cluster Ø Spark Ø 5 c3.8xlarge VMs Ø A larger cluster yields bexer performance # Users 0.5M 1M 2M 5M 10M # Edges ~8M ~16M ~32M ~80M ~160M Execu5on 5me 288 sec 669 sec 1767 sec 8049 sec 7.7 hours 47

48 Rejecto can handle multimillion- user social graphs Performance on an EC2 cluster Ø Spark Execution time grows gracefully with the graph size Ø 5 c3.8xlarge VMs Ø A larger cluster yields bexer performance # Users 0.5M 1M 2M 5M 10M # Edges ~8M ~16M ~32M ~80M ~160M Execu5on 5me 288 sec 669 sec 1767 sec 8049 sec 7.7 hours 48

49 Conclusion Rejecto: uncovers friend spammers using social rejections Ø Immune to axack strategies Ø Efficient Ø Scalable 49