Landing page for everything solar. Here’s a link home to get started.
Introducing the first release of a new public benchmark dataset of solar image from the Solar Dynamics Observatory (SDO) and solar event metadata from the SDO Feature Finding Team (FFT). It covers the first six months of 2012 and contains over 15,000 images and 24,000 region-based event labels. Also included is ten pre-computed image parameters […]
ImageFARMER is a FrAmework for the cReation of large-scale content-based iMage rEtRieval systems http://www.imagefarmer.org/
Spatio-temporal co-occuring patterns represent subsets of event types that occur together in both time and space. This project provides a general framework to mining spatio-temporal co-occurring patterns for spatio-temporal events.
Modern databases and data mining tasks are often faced with costly data retrieval tasks in very high dimensional data. We present an open source implementation of the multi-dimensional indexing method iDistance and our extensions. It includes many customizable algorithm options and a set of basic test files (and dataset) with instructions to use and install […]
Managing and accommodating the evolution of database schema (the metadata describing the structure of a database) poses a number of interesting problems. Some of these problems are particularly acute in Online Transaction Processing (OLTP) databases that serve as the data store for large, extremely active data processing systems, especially in systems that use a Software […]
The solar CBIR system is the first ever Content-Based Image Retrieval system specifically suited for solar image archives. Our Feature Finding Team (FFT) — part of NASA’s SDO Mission — is currently creating this system and a publicly accessible website to interact with it. Users will be able to use images of interest to search […]
Welcome to Georgia State University’s Data Mining Lab website. More to come soon!