The real-time detection of anomalous phenomena on streaming data has become increasingly important for applications ranging from fraud detection, financial analysis to traffic management. In these streaming applications, often a large number of similar continuous outlier detection queries are executed concurrently. In the light of the high algorithmic complexity of detecting and maintaining outlier patterns for different parameter settings independently, we propose a shared execution methodology called SOP that handles a large batch of requests with diverse pattern configurations.
First, our systematic analysis reveals opportunities for maximum resource sharing by leveraging commonalities among outlier detection queries. For that, we introduce a sharing strategy that integrates all computation results into one compact data structure. It leverages temporal relationships among stream data points to prioritize the probing process. Second, this work is the first to consider predicate constraints in the outlier detection context. By distinguishing between target and scope constraints, customized fragment sharing and block selection strategies can be effectively applied to maximize the efficiency of system resource utilization. Our experimental studies utilizing real stream data demonstrate that our approach performs 3 orders of magnitude faster than the start-of-the-art and scales to 1000s of queries.
Karen Works' research has focussed on the development of novel technology for handling stream data with multiple levels of importance to an organization, and in particular, its multi-tiered priority-based query processing in the face of limited resources. Karen's contributions include but are not limited to the following core innovations:
We wish Dr. Karen Works best of luck in her professional career as Tenure-Track Professor of Computer Science at Westfield State University!
Di Wang's dissertation research has focussed on the development of several critical innovations within the context of complex event processing over high-volume data streams to further emerging applications ranging from on-line financial transactions, RFID based supply chain management to real-time object monitoring. In particular, her contributions include innovations and publications in top venues on the following topics: a. For applications which require access to both streaming and stored data, she has introduced an active complex CEP model with clear semantics and efficient scheduler algorithms in the face of concurrent access and failures. b. When deployed in a sensitive environment such as health care, she has proposed event-suppression technology critical for mitigating possible privacy leaks within the context of complex event processing systems. c. For high-performance inferencing of probabilistic identification of events with possible missing identifiers, her work not only provides a graphical model to capture this inference problem but she also designed general system optimizations that speed up existing inference strategies on streams up to 15 fold.
In addition, she was a key architect and developer of the HyReminder Web Application for employing CEP technology to track health care workers's activities at the UMASS Memorial Hospital. This software, currently deployed at UMASS ICUs, has undergone a clinical trial - showing clear positive indicators of the effectiveness of such electronic reminder technology. Results of this health care trial have been submitted to a health care meeting.
We thank everyone who was able to attend Di's presentation today and to lend Di support. In particular we would like to thank the committee members Prof. Dougherty, Prof. Eltabakh and the external committee member Dr. Badrish Chandramouli from Microsoft Research Labs for their time and effort in guiding Di through her PhD studies. We also thank the DSRG lab members for listening to Di's research, sharing ideas, and generally supporting each other over these years.
Now we wish Di best of luck in her professional career starting at BING! at Microsoft Corp.effective immediately.
Congratulations to Venky on having successfully conducted high-quality and innovative research, which has been published in top venues, including ICDE, Information Systems Journal, IDAR, and others, and several very well-received software demonstrations of core technologies in ACM SIGMOD. Venky's dissertation research falls in the area of big-data analytics and multi-criteria preference systems. His dissertation is entitled "Supporting Multi-Criteria Decision Support Queries over Disparate Data Sources". Given the exponential growth of information, providing services to help analysts, businesses and users alike to extract value from data is imperative for staying ahead and meeting one's information needs. In this context, Venky has designed a suite of innovative techniques and corresponding software technologies that tackle open problems in support of multi-dimensional preference (skyline) queries, enabling users to quickly grasp their prefered choices from a huge data store.
Venky has started his professional career at the Greenplum startup (now, an EMC company) in California, and is enjoying every day of it. He is getting his hands deep into the guts of a commercial query optimizer for large-scale distributed compute platforms - helping to build it from the grounds up to meet the BigData buzz. We wish him the very best success and fun in his future professional career in computing!
We also would like to thank everyone who was able to attend Venky's defense yesterday and lend their support to him. We thank the CS department, all faculty, the office and computing staff, for providing an amazingly nuturing environment in which Venky could mature into an accomplished researcher and Computer Scientist. It sure was a pleasure yesterday seeing Venky shine in his accomplishments -- he has come a long way, and I am proud of him. In particular, we thank the committee members Prof. Dan Dougherty, Prof. Murali Mani and Dr. Haixun Wang (Microsoft Research Asia) for their time, effort and extremely valuable feedback on Venky's work. Their help in guiding Venky is very much appreciated.
The committee has accepted her work subject to minor revisions, which Mo plans to apply to the manuscript in the following weeks. Congratulations to soon-to-be Dr. Liu! Congratulations to Mo on having conducted high-quality research, which has been published in top venues in the database field, including SIGMOD, ICDE, and others.
Mo's dissertation research falls in the area of Complex Event Processing on Data Streams. Specifically, her dissertation entitled "Extending Event Sequence Processing: New Models and Optimization Techniques" includes the design, development and evaluation of several techniques at the core of an E-Analytic system to achieve efficient, scalable and robust methods for in-memory multi-dimensional nested pattern analysis over high-speed event streams.
We would like to thank everyone who was able to attend Mo's defense and lend their support to Mo. We thank the CS department for providing a nuturing environment in which Mo Liu could mature into an accomplished Computer Scientist. In particular, we thank the committee members Prof. Dan Dougherty, Prof. Yanlei Diao, University of Massachusetts Amherst; Prof. Murali Mani, University of Michigan, Flint; and Prof. Ismail Ari, Ozyegin University, Turkey for their time and valuable feedback on Moi's work. In particular, Mo would like to extend a special thank you to Prof. Dan Dougherty, who has spent countless hours in helping Mo to explore the world of CEP language design, semantics and optimization. The committee's help in guiding Mo to make her work of the utmost quality is much appreciated.
Lastly, Mo Liu has started her professional career at Sybase, Inc, an SAP Company, in California. We wish her a fulfilling career complete with interesting challenges and both success and fun going forward, where ever life may take her.