Incrementally Maintaining Classification using an RDBMS

The proliferation of imprecise data has motivated both researchers and the database industry to push statistical techniques into relational database management systems (RDBMSes). We study strategies to maintain model-based views for a popular statistical technique, classification, inside an RDBMS in the presence of updates (to the set of training examples). We make three technical contributions: (1) A strategy that incrementally maintains classification inside an RDBMS. (2) An analysis of the above algorithm that shows that our algorithm is optimal among all deterministic algorithms (and asymptotically within a factor of 2 of a non-deterministic optimal strategy). (3) A novel hybrid-architecture based on the technical ideas that underlie the above algorithm which allows us to store only a fraction of the entities in memory. We apply our techniques to text processing, and we demonstrate that our algorithms provide an order of magnitude improvement over non-incremental approaches to classification on a variety of data sets, such as the Citeseer and DBLife.

[1]  Jiawei Han,et al.  Maintenance of discovered association rules in large databases: an incremental updating technique , 1996, Proceedings of the Twelfth International Conference on Data Engineering.

[2]  Surajit Chaudhuri,et al.  Efficient evaluation of queries with mining predicates , 2002, Proceedings 18th International Conference on Data Engineering.

[3]  Lise Getoor,et al.  PrDB: managing and exploiting rich correlations in probabilistic databases , 2009, The VLDB Journal.

[4]  Benjamin Recht,et al.  Random Features for Large-Scale Kernel Machines , 2007, NIPS.

[5]  Jayant Madhavan,et al.  Reference reconciliation in complex information spaces , 2005, SIGMOD '05.

[6]  Manfred K. Warmuth,et al.  The Weighted Majority Algorithm , 1994, Inf. Comput..

[7]  Koby Crammer,et al.  Online Passive-Aggressive Algorithms , 2003, J. Mach. Learn. Res..

[8]  Anna R. Karlin,et al.  Competitive randomized algorithms for non-uniform problems , 1990, SODA '90.

[9]  Rakesh Agrawal,et al.  SPRINT: A Scalable Parallel Classifier for Data Mining , 1996, VLDB.

[10]  Surajit Chaudhuri,et al.  Extracting predicates from mining models for efficient query evaluation , 2004, TODS.

[11]  Daniel S. Weld,et al.  Automatically refining the wikipedia infobox ontology , 2008, WWW.

[12]  Daphne Koller,et al.  Support Vector Machine Active Learning with Applications to Text Classification , 2000, J. Mach. Learn. Res..

[13]  Samuel Madden,et al.  MauveDB: supporting model-based user views in database systems , 2006, SIGMOD Conference.

[14]  Peter J. Haas,et al.  MCDB: a monte carlo approach to managing uncertain data , 2008, SIGMOD Conference.

[15]  Gert Cauwenberghs,et al.  Incremental and Decremental Support Vector Machine Learning , 2000, NIPS.

[16]  Klaus-Robert Müller,et al.  Incremental Support Vector Learning: Analysis, Implementation and Applications , 2006, J. Mach. Learn. Res..

[17]  Anuradha Bhamidipaty,et al.  Interactive deduplication using active learning , 2002, KDD.

[18]  Marcos M. Campos,et al.  SVM in Oracle Database 10g: Removing the Barriers to Widespread Adoption of Support Vector Machines , 2005, VLDB.

[19]  Srinivasan Parthasarathy,et al.  Efficiently Mining Approximate Models of Associations in Evolving Databases , 2002, PKDD.

[20]  W. Rudin.  Principles of mathematical analysis , 1964 .

[21]  Jennifer Widom,et al.  Databases with uncertainty and lineage , 2008, The VLDB Journal.

[22]  Larry Wasserman,et al.  All of Nonparametric Statistics (Springer Texts in Statistics) , 2006 .

[23]  Ali R. Hurson,et al.  TF-ICF: A New Term Weighting Scheme for Clustering Dynamic Data Streams , 2006, 2006 5th International Conference on Machine Learning and Applications (ICMLA'06).

[24]  Frank Wm. Tompa,et al.  Efficiently updating materialized views , 1986, SIGMOD '86.

[25]  Dan Olteanu,et al.  Fast and Simple Relational Processing of Uncertain Data , 2007, 2008 IEEE 24th International Conference on Data Engineering.

[26]  Thorsten Joachims,et al.  Making large scale SVM learning practical , 1998 .

[27]  Jörg Kindermann,et al.  Text Categorization with Support Vector Machines. How to Represent Texts in Input Space? , 2002, Machine Learning.

[28]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[29]  Srinivasan Parthasarathy,et al.  Mining Frequent Itemsets in Evolving Databases , 2002, SDM.

[30]  Amol Deshpande,et al.  Online Filtering, Smoothing and Probabilistic Modeling of Streaming data , 2008, 2008 IEEE 24th International Conference on Data Engineering.

[31]  Petra Perner,et al.  Data Mining - Concepts and Techniques , 2002, Künstliche Intell..

[32]  Yoram Singer,et al.  Pegasos: primal estimated sub-gradient solver for SVM , 2011, Math. Program..

[33]  Dan Suciu,et al.  Efficient query evaluation on probabilistic databases , 2004, The VLDB Journal.

[34]  Christopher Ré,et al.  Event queries on correlated probabilistic streams , 2008, SIGMOD Conference.