Abstract: Recently, sequence anomaly detection has been widely used in many fields. Sequence data in these fields are usually multi-dimensional over the data stream. It is a challenge to design an anomaly detection method for a multi-dimensional sequence over the data stream to satisfy the requirements of accuracy and high speed. It is because: (1) Redundant dimensions in sequence data and large state space lead to a poor ability for sequence modeling; (2) anomaly detection cannot adapt to the high-speed nature of the data stream, especially when concept drift occurs, and it will reduce the detection rate. On one hand, most existing methods of sequence anomaly detection focus on the single-dimension sequence. On the other hand, some studies concerning multi-dimensional sequence concentrate mainly on the static database rather than the data stream. To improve the performance of anomaly detection for a multi-dimensional sequence over the data stream, we propose a novel unsupervised fast and accurate anomaly detection (FAAD) method which includes three algorithms. First, a method called “information calculation and minimum spanning tree cluster” is adopted to reduce redundant dimensions. Second, to speed up model construction and ensure the detection rate for the sequence over the data stream, we propose a method called “random sampling and subsequence partitioning based on the index probabilistic suffix tree.” Last, the method called 𠇊nomaly buffer based on model dynamic adjustment” dramatically reduces the effects of concept drift in the data stream. FAAD is implemented on the streaming platform Storm to detect multi-dimensional log audit data. Compared with the existing anomaly detection methods, FAAD has a good performance in detection rate and speed without being affected by concept drift.

FAAD：一种无监督快速准确的数据流上多维序列异常检测方法

[1]Bao H, Wang YJ, 2016. A C-SVM based anomaly detection method for multi-dimensional sequence over data stream. Proc IEEE 22^nd Int Conf on Parallel and Distributed Systems, p.948-955.

[2]Box GE, Jenkins GM, Reinsel GC, et al., 2015. Time Series Analysis: Forecasting and Control. John Wiley & Sons, Hoboken, USA.

[3]Budalakoti S, Srivastava AN, Akella R, et al., 2006. Anomaly Detection in Large Sets of High-Dimensional Symbol Sequences. TM-2006-214553, NASA Ames Research Center, USA.

[4]Budalakoti S, Srivastava AN, Otey ME, 2009. Anomaly detection and diagnosis algorithms for discrete symbol sequences with applications to airline safety. IEEE Trans Syst Man Cybern C, 39(1):101-113.

[5]Carlin BP, Louis TA, 2000. Bayes and Empirical Bayes Methods for Data Analysis (2^nd Ed.). Chapman & Hall/linebreak CRC Press, Boca Raton, FL, USA.

[6]Chandola V, Mithal V, Kumar V, 2008. Comparative evaluation of anomaly detection techniques for sequence data. Proc 8^th IEEE Int Conf on Data Mining, p.743-748.

[7]Chandola V, Banerjee A, Kumar V, 2009. Anomaly detection: a survey. ACM Comput Surv, 41(3), Article 15.

[8]Chandola V, Banerjee A, Kumar V, 2012. Anomaly detection for discrete sequences: a survey. IEEE Trans Knowl Data Eng, 24(5):823-839.

[9]Dani MC, Freixo C, Jollois FX, et al., 2015. Unsupervised anomaly detection for aircraft condition monitoring system. Proc IEEE Aerospace Conf, p.1-7.

[10]Esposito F, di Mauro N, Basile TMA, et al., 2008. Multi-dimensional relational sequence mining. Fundam Inform, 89(1):23-43.

[11]Hall MA, 2000. Correlation-based feature selection for discrete and numeric class machine learning. Proc 17^th Int Conf on Machine Learning, p.359-366.

[12]Jin Y, Zuo WL, 2007. Multi-dimensional concept lattice and incremental discovery of multi-dimensional sequential patterns. J Comput Res Dev, 44(11):1816-1824 (in Chinese).

[13]Kaufman L, Rousseeuw PJ, 2009. Finding Groups in Data: an Introduction to Cluster Analysis. John Wiley & Sons, New York, USA.

[14]Keogh E, Chakrabarti K, Pazzani M, et al., 2001. Dimensionality reduction for fast similarity search in large time series databases. Knowl Inform Syst, 3(3):263-286.

[15]Kponyo JJ, Kuang YJ, Zhang EZ, et al., 2013. VANET cluster-on-demand minimum spanning tree (MST) prim clustering algorithm. Proc Int Conf on Computational Problem-Solving, p.101-104.

[16]Lane T, 1998. Machine Learning Techniques for the Domain of Anomaly Detection for Computer Security. Purdue University, Indiana, USA.

[17]Lee CH, 2015. A multi-phase approach for classifying multi-dimensional sequence data. Intell Data Anal, 19(3):547-561.

[18]Li C, Tian XG, Xiao X, et al., 2012. Anomaly detection of user behavior based on shell commands and co-occurrence matrix. J Comput Res Dev, 49(9):1982-1990 (in Chinese).

[19]Li XY, Wang YJ, Li XL, et al., 2014. Parallelizing skyline queries over uncertain data streams with sliding window partitioning and grid index. Knowl Inform Syst, 41(2):277-309.

[20]Parveen P, Mcdaniel N, Weger Z, et al., 2013. Evolving insider threat detection stream mining perspective. Int J Artif Intell Tools, 22(5):1360013.

[21]Qian Q, Wu JL, Zhu W, et al., 2012. Improved edit distance method for system call anomaly detection. Proc IEEE 12^th Int Conf on Computer and Information Technology, p.1097-1102.

[22]Ron DN, Singer Y, Tishby N, 1994. Learning probabilistic automata with variable memory length. Proc 7^th Annual Conf on Computational Learning Theory, p.35-46.

[23]Sarhrouni E, Hammouch A, Aboutajdine D, 2012. Application of symmetric uncertainty and mutual information to dimensionality reduction and classification of hyperspectral images. Int J Eng Technol, 4(5):268-276.

[24]Shu XK, Yao DF, Ryder BG, 2015. A formal framework for program anomaly detection. Proc 18^th Int Symp Research in Attacks, Intrusions, and Defenses, p.270-292.

[25]Tandon G, Chan P, 2003. Learning rules from system call arguments and sequences for anomaly detection. Proc ICDM Workshop on Data Mining for Computer Security, p.20-29.

[26]Wang Y, Ma X, 2015. A general scalable and elastic content-based publish/subscribe service. IEEE Trans Parall Distr Syst, 26(8):2100-2113.

[27]Wang YJ, Li S, 2006. Research and performance evaluation of data replication technology in distributed storage systems. Comput Math Appl, 51(11):1625-1632.

[28]Wang YJ, Li XY, Li XL, et al., 2013. A survey of queries over uncertain data. Knowl Inform Syst, 37(3):485-530.

[29]Wang YJ, Pei X, Ma X, et al., 2018. TA-update: an adaptive update scheme with tree-structured transmission in erasure-coded storage systems. IEEE Trans Parall Distr Syst, 29(8):1893-1906.

[30]Xianyu JC, Rasouli S, Timmermans H, 2017. Analysis of variability in multi-day GPS imputed activity-travel diaries using multi-dimensional sequence alignment and panel effects regression models. Transportation, 44(3):533-553.

[31]Xiong TK, Wang SR, Jiang QS, et al., 2011. A new Markov model for clustering categorical sequences. Proc IEEE 11^th Int Conf on Data Mining, p.854-863.

[32]Yamanishi K, Maruyama Y, 2005. Dynamic syslog mining for network failure monitoring. Proc 11^th ACM SIGKDD Int Conf on Knowledge Discovery in Data Mining, p.499-508.

[33]Yang J, Wang W, 2003. CLUSEQ: efficient and effective sequence clustering. Proc 19^th Int Conf on Data Engineering, p.101-112.

[34]Yu L, Liu H, 2003. Feature selection for high-dimensional data: a fast correlation-based filter solution. Proc 20^th Int Conf on Machine Learning, p.856-863.