p169-metwally - Using Association Rules for Fraud Detection...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Using Association Rules for Fraud Detection in Web Advertising Networks Ahmed Metwally Divyakant Agrawal Amr El Abbadi Department of Computer Science University of California, Santa Barbara Santa Barbara CA 93106 { metwally, agrawal, amr } @cs.ucsb.edu Abstract Discovering associations between elements oc- curring in a stream is applicable in numerous applications, including predictive caching and fraud detection. These applications require a new model of association between pairs of el- ements in streams. We develop an algorithm, Streaming-Rules , to report association rules with tight guarantees on errors, using limited processing per element, and minimal space. The modular design of Streaming-Rules allows for integration with current stream manage- ment systems, since it employs existing tech- niques for finding frequent elements. The pre- sentation emphasizes the applicability of the algorithm to fraud detection in advertising networks. Such fraud instances have not been successfully detected by current techniques. Our experiments on synthetic data demon- strate scalability and eciency. On real data, potential fraud was discovered. 1 Introduction Recently, online monitoring of data streams has emerged as an important data management problem. This research topic has its foundations and appli- cations in many domains, including databases, data mining, algorithms, networking, theory, and statistics. However, new challenges have emerged. Due to their vast sizes, some stream types should be mined fast before being deleted forever. In general, the alpha- This work was supported in part by NSF under grants EIA 00-80134, NSF 02-09112, and CNF 04-23336. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the VLDB copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Very Large Data Base Endowment. To copy otherwise, or to republish, requires a fee and/or special permission from the Endowment. Proceedings of the 31st VLDB Conference, Trondheim, Norway, 2005 bet is too large to keep exact information for all ele- ments. Conventional database, and mining techniques are deemed impractical in this setting. In this paper, we develop the notion of association rules in streams of elements. To the best of our knowl- edge, this problem has not been addressed before. The data model in recent dependency detection research, [4, 20, 22], is that of the classical dependency detec- tion mining [1, 2], with the exception that the tech- niques are applied to data streams, rather than stored data. That is, the data model is that of a stream of customers transactions with a large number of cus- tomers and a limited number of items per transaction....
View Full Document

Page1 / 12

p169-metwally - Using Association Rules for Fraud Detection...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online