DC: Small: Stream Clustering Algorithms in Mixed Domains with Soft Two-way Semi-Supervision

点击翻译按钮获取中文摘要

One way to form a model of massive data sets is to use clustering techniques that summarize the data by several cluster representatives. However, clustering huge data sets is a very challenging problem whose difficulty increases further when the data is dynamic. We are developing scalable and robust stream summarization methods to provide a concise summary of huge multi-dimensional data streams that keep track of each discovered cluster or component of the summary through time, and that store only milestones corresponding to the occurrence of significant changes in these cluster representatives. Moreover to handle possibly diverse data formats and different sources of data, we are using a semi-supervised framework for (i) combining diverse representations of the data, in particular where data comes from different sources, some of which may be unreliable or uncertain; and (ii) exploiting optional external concept set labels to guide the clustering of the main data set in its original domain.Our methods have tremendous impact on applications that deal with streaming data in general, and more specifically on monitoring data streams in real-life dynamic settings. For example, as more and more everyday activities move online, network and Web data has been increasing at a rapid pace that precludes standard and classical data analysis methods, and call instead for real time analysis. The same can be said about the deluge of data that is being or about to be generated by new and future sensor networks, astronomical observatories, and missions in space. Thus, new research efforts and paradigms are needed and will have a strong impact on our ability to digest and make sense of this data.