Nuclear Activation Analysis System (NAAS)

Nuclear Activation Analysis System (NAAS) is designed for searching radiation and isotopes information in nuclear reactions, estimating the radioactivity before irradiation starts, and Nuclear Activation Analysis (including NAA and PAA). Our samples are mostly related to agriculture, such as seeds, soil, fertilizer, grain, vegetable, etc. However, the analysis system can be applied to samples in other disciplines as well.

The database was retrieved from the Lund/LBNL Nuclear Data Search website, the Experimental Nuclear Reaction Data (EXFOR) website, the book “Photon Activation Analysis” by Dr. Christian Segebade, and “Tables for Analytical Methods at MURR: NAA, XRF and ICP-MS” by Dr. Michael Glascock. Some data and the corresponding formats are revised in order to better serve the nuclear radiation research at the Radiation Data Mining Laboratory in the South Carolina State University.

This research is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, Evans-Allen project number SCX-312-06-16 under the leadership of 1890 Research & Extension at the South Carolina State University.

Any opinions, findings, conclusions, or recommendations expressed at this website are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture.

To visit the website of NAAS, please click here.




TDMiner is a temporal data mining tool that mines frequent episodes from the long sequence of events. The program is written in two versions (Java and C++) and provides a suite of mining techniques for different types of frequent episodes. It also has tools for visualizing the results obtained from mining. The current version of TDMiner is inherited from the Gminer program of the General Motor, which was used to search the frequency episodes in automobile assembling lines in Michigan and online catalog of (Patnaik 2012). The original code of the program can be obtained at the Github website.


TDMiner code at Github

#TDMiner C++: Frequent episode mining with inter-event gap constraints.

Usage: tdminer [ht:i:m:sxcdz:o:f:] eventsfile outcomefile [min_supp]

 eventsfile	    file, that contains the event stream
 outcomefile	    file to write the outcome
 min_supp	    support threshold (default 0.0100)

Basic options:
 -h			Gives this help display.
 -t <num>		Specifies the number of pthreads (default 1).
 -i <intervalfile>	Specifies the interval file (default ivl.txt).
			See example ivl.txt for format
 -m <num>		Specifies the maximum level up to which mining is done.
			(default -1 i.e. no limit). Level corresponds to episode size.
 -s			Specifies minimum support as a ratio of #occurrences to total time span.
			By default minimum support is the ratio of #occurrences
			to total number of events in the data.
 -x			Turn off pre-count pruning heuristics.
 -c			Takes the first column as customer id
			and prevents patterns from crossing over across customers..
 -d			Enables duration of events. Reads the last column as end time of an event
 -z <num>		Set the list size for storing time stamps.
 -o <epsfile>		Counts the episodes specified in <epsfile>.
			The result of counting is stored in <outcomefile>.
 -f <occurfile>		Specifies the file into which episode occurrences are written.
			This works only in conjunction with -o
			The occurrences are output as <episode_index> : <list of time stamps>.

Mining Example:
$ tdminer sample_stream.csv test_episodes.txt 0.01

Counting example:
$ tdminer -o eps.txt -f occurrences.txt sample_stream.csv episodes-out.txt
"eps.txt" contains the episodes to count. Note currently only episode of 
the same size are supported.