Extensions

Use of Extensions

Add the extension jar file to the Java classpath

java -cp extension.jar:moa.jar -javaagent:sizeofag.jar moa.gui.GUI

New Extensions

  • KUE Kappa Updated Ensemble presented in A. Cano and B. Krawczyk. Kappa Updated Ensemble for Drifting Data Stream Mining. Machine Learning, DOI: 10.1007/s10994-019-05840-z, 2019. Website
  • Multi-label Punitive kNN with Self-Adjusting Memory presented in M. Roseberry, B. Krawczyk, and A. Cano. Multi-label Punitive kNN with Self-Adjusting Memory for Drifting Data Streams. ACM Transactions on Knowledge Discovery from Data, 2019. Website
  • G-eRules is a streaming classification algorithm based in rules. The G stands for the tuse of the Gaussian distribution and the algorithm incorporates a new method for continuous attribute. It was presented in Thien Le, Frederic Stahl, Joao Gomes, Mohamed Medhat Gaber, Giuseppe Di Fatta (2014) Computationally Efficient Rule-Based Classification for Continuous Streaming Data. In Proceedings of the Thirty-Fourth SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence. Website
  • MOAReduction is an extension for MOA, which allows us to perform data reduction techniques on streams of data (w/o drift). It includes several reduction methods for different tasks, such as: discretization, instance selection, and feature selection, presented in https://dx.doi.org/10.1016/j.neucom.2017.01.078 Sergio Ramírez-Gallego, Bartosz Krawczyk, Salvador García, Michał Woźniak, Francisco Herrera, A survey on Data Preprocessing for Data Stream Mining: Current status and future directions, Neurocomputing, Website
  • Social Adaptive Ensemble 2 (SAE2), Scale-free Network Classifier
    (SFNClassifier) and the Social Network Clusterer Stream (SNCStream).
    MOA social based algorithms by Fabrício Enembreck, Heitor Murilo Gomes, and Jean Paul Barddal. Website

 

Available Extensions

Frequent Pattern Mining

  • MOA-IncMine IncMine is an extension for MOA to compute Frequent Closed Itemsets from data streams. It implements the method proposed by James Cheng, Yiping Ke and Wilfred Ng “Maintaining frequent closed itemsets over a sliding window” Journal of Intelligent Information Systems, 2008, Volume 31, Number 3, Pages 191-215, using a version of the CHARM algorithm, proposed by Zaki et al., to compute frequent closed itemsets over a batch of transactions. Website
  • MOA-AdaGraphMiner AdaGraphMiner is a framework for studying graph pattern mining on time-varying streams. It contains three new methods for mining frequent closed subgraphs. All methods work on coresets of closed subgraphs, compressed representations of graph sets, and maintain these sets in a batch-incremental manner, but use different approaches to address potential concept drift. Website
  • MOA-Moment Moment is a closed frequent itemset miner over a stream sliding window. Implemented by Maciek Jarka (www.admire-project.eu). It was presented in: Yun Chi, Haixun Wang, Philip S. Yu, Richard R. Muntz: Moment: Maintaining Closed Frequent Itemsets over a Stream Sliding Window. ICDM 2004: 59-66 Download

Classification and Regression

  • IBLStreams (Instance Based Learner on Streams): is an instance-based learning algorithm for classification and regression problems on data streams. By Ammar Shaker, Eyke Hüllermeier and Jürgen Beringer. Website
  • MOA-TweetReader MOA-TweetReader reads and converts tweets from the Twitter Streaming API to MOA instances, so that it is possible to perform data stream mining with them. Website
  • Classifiers & Drift Detection Methods This extension provides several published ensemble classifiers (DWM, RCD, Learn++.NSE, EB), concept drift detectors (ECDD, PHT, Paired Learners), and artificial data sets (Sine and Mixed). Website
  • MODL split criterion and GK class summary This new split criterion is based on the MODL approach from Marc Boullé. The GK class summary is based on Greenwald and Khanna quantile summary but in this version class counts are included in each tuple in the summary. Website
  • iOVFDT (Incrementally Optimized Very Fast Decision Tree): New decision tree proposed by Hang Yang and Simon Fong in: Incrementally Optimized Decision Tree for Noisy Big Data, The 1st ACM SIGKDD Workshop on Big Data Mining (BigMine) Beijing, P.R. China Aug, 2012
    Website
  • Anytime Nearest Neighbor Classifier Implementation by Liang Liu and Reuben Bell of the anytime classifier presented in: Shieh, J. and Keogh, E. (2010). Polishing the Right Apple: Anytime Classification Also Benefits Data Streams with Constant Arrival Times. ICDM’10
    Website
  • Social Adaptive Ensemble 2 (SAE2), Scale-free Network Classifier
    (SFNClassifier) and the Social Network Clusterer Stream (SNCStream).
    MOA Social based algorithms by Fabrício Enembreck, Heitor Murilo Gomes, and Jean Paul Barddal. Website

Stream Sentiment Analysis

  • Framework for Sentiment Analysis of a Stream of Texts (2012 Harvest Project) This project aims at building an online, real-time system able to analyze an incoming stream of text and visualize its main characteristics using a minimal desktop application. The software is designed to work in a “stream setting” with minimal memory/space requirements. Website

Feature Processing

  • MOAReduction is an extension for MOA, which allows us to perform data reduction techniques on streams of data (w/o drift). It includes several reduction methods for different tasks, such as: discretization, instance selection, and feature selection, presented in https://dx.doi.org/10.1016/j.neucom.2017.01.078 Sergio Ramírez-Gallego, Bartosz Krawczyk, Salvador García, Michał Woźniak, Francisco Herrera, A survey on Data Preprocessing for Data Stream Mining: Current status and future directions, Neurocomputing, Website

Android

  • MOA for Android. Software to make MOA (Massive Online Analysis) usable as part of Android application. Website