Event mining encompasses techniques for automatically and efficiently extracting valuable knowledge from historical event/log data. The field, therefore, plays an important role in data-driven system management. Event Mining: Algorithms and Applications presents state-of-the-art event mining approaches and applications with a focus on computing system management.
The book first explains how to transform log data in disparate formats and contents into a canonical form as well as how to optimize system monitoring. It then shows how to extract useful knowledge from data. It describes intelligent and efficient methods and algorithms to perform data-driven pattern discovery and problem determination for managing complex systems. The book also discusses data-driven approaches for the detailed diagnosis of a system issue and addresses the application of event summarization in Twitter messages (tweets).
Understanding the interdisciplinary field of event mining can be challenging as it requires familiarity with several research areas and the relevant literature is scattered in diverse publications. This book makes it easier to explore the field by providing both a good starting point for readers not familiar with the topics and a comprehensive reference for those already working in this area.
Preface |
|
xiii | |
|
|
xv | |
|
|
xxi | |
Editor |
|
xxiii | |
Contributors |
|
xxv | |
|
|
1 | (10) |
|
|
1.1 Data-Driven System Management |
|
|
1 | (4) |
|
|
5 | (1) |
|
|
6 | (3) |
|
|
9 | (2) |
|
I Event Generation and System Monitoring |
|
|
11 | (58) |
|
2 Event Generation: From Logs to Events |
|
|
13 | (36) |
|
|
|
|
14 | (3) |
|
|
17 | (2) |
|
2.3 Log Message Classification |
|
|
19 | (2) |
|
2.4 Log Message Clustering |
|
|
21 | (3) |
|
2.5 Tree-Structure-Based Clustering |
|
|
24 | (7) |
|
2.6 Message-Signature-Based Event Generation |
|
|
31 | (15) |
|
|
46 | (1) |
|
|
46 | (3) |
|
3 Optimizing System Monitoring Configurations |
|
|
49 | (20) |
|
|
|
|
49 | (1) |
|
|
50 | (4) |
|
3.3 Eliminating False Positives |
|
|
54 | (3) |
|
3.4 Eliminating False Negatives |
|
|
57 | (3) |
|
|
60 | (6) |
|
|
66 | (1) |
|
|
66 | (3) |
|
II Pattern Discovery and Summarization |
|
|
69 | (124) |
|
|
71 | (52) |
|
|
|
|
72 | (2) |
|
|
74 | (6) |
|
4.3 Fully Dependent Pattern |
|
|
80 | (2) |
|
4.4 Partially Periodic Dependent Pattern |
|
|
82 | (6) |
|
4.5 Mutually Dependent Pattern |
|
|
88 | (3) |
|
|
91 | (5) |
|
|
96 | (2) |
|
|
98 | (3) |
|
|
101 | (1) |
|
4.10 Correlated Pattern between Time Series and Event |
|
|
102 | (2) |
|
|
104 | (16) |
|
|
120 | (1) |
|
|
120 | (3) |
|
|
123 | (28) |
|
|
|
|
|
123 | (2) |
|
5.2 Non-Parametric Method |
|
|
125 | (7) |
|
|
132 | (8) |
|
|
140 | (9) |
|
|
149 | (1) |
|
|
149 | (2) |
|
6 Log Event Summarization |
|
|
151 | (42) |
|
|
|
|
152 | (5) |
|
6.2 Summarizing with Frequency Change |
|
|
157 | (8) |
|
6.3 Summarizing with Temporal Dynamics |
|
|
165 | (14) |
|
6.4 Facilitating the Summarization Tasks |
|
|
179 | (12) |
|
|
191 | (1) |
|
|
192 | (1) |
|
|
193 | (90) |
|
7 Data-Driven Applications in System Management |
|
|
195 | (62) |
|
|
|
|
|
|
196 | (5) |
|
7.2 Searching Similar Sequential Textual Event Segments |
|
|
201 | (22) |
|
7.3 Hierarchical Multi-Label Ticket Classification |
|
|
223 | (13) |
|
7.4 Ticket Resolution Recommendation |
|
|
236 | (17) |
|
|
253 | (2) |
|
|
255 | (2) |
|
8 Social Media Event Summarization Using Twitter Streams |
|
|
257 | (26) |
|
|
|
|
258 | (2) |
|
|
260 | (2) |
|
8.3 Tweet Context Analysis |
|
|
262 | (3) |
|
8.4 Sub-Event Detection Methods |
|
|
265 | (5) |
|
8.5 Multi-Tweet Summarization |
|
|
270 | (1) |
|
|
271 | (10) |
|
8.7 Conclusion and Future Work |
|
|
281 | (1) |
|
|
281 | (2) |
Bibliography |
|
283 | (22) |
Index |
|
305 | |
Dr. Tao Li is a professor and Graduate Program Director in the School of Computing and Information Sciences at Florida International University (FIU) and a professor in the School of Computer Science at Nanjing University of Posts and Telecommunication. He is on the editorial boards of ACM Transactions on Knowledge Discovery from Data, IEEE Transactions on Knowledge and Data Engineering, and Knowledge and Information System Journal. He has received numerous honors, including an NSF CAREER Award, IBM Faculty Research Awards, an FIU Excellence in Research and Creativities Award, and IBM Scalable Data Analytics Innovation Award and Mentorship Awards. His research interests are in data mining, information retrieval, and computing system management. He received a PhD in computer science from the University of Rochester.