Efficient sequential pattern mining using segmented event task optimizer in temporal database

Abstract

Data mining algorithms are designed as a mechanism to ascertain newlinerelevant patterns from extensive amounts of data Temporal database which newlinemany applications are the foremost data mining task over the mining newlinesequential data The conventional association rule mining algorithm extracts newlinethe rules based on the attributes which are found often in a data set A pattern newlinewith an ordered list of events and tasks repeats for a particular time of newlineinterval To identify the patterns which appear in a sequential data format in newlinetemporal database most work on sequential mining which uses the support newlinevalue as the metric to identify and extract the patterns from the other patterns newlineIn order to effectively extract the patterns from a temporal database newlinewe have to extract only valuable have to be extracted which are patterns newlineincluded in sequence data by skipping the patterns which are of no use newlineA temporal database for mining sequential type of data collects a sequence of newlinedata values gathered at regular interval of time In this work a more novel newlinetechnique is proposed using a segmentation method over temporal database newlineby way of performing random segmentation model detecting potential cycles newlineand finally applying the event task optimizer to extract the more frequent newlinepatterns observed for a particular interval of time newlineInitially an efficient Segmented Sequential Pattern Mining newlinealgorithm for mining sequential data is presented using segmentation method newlineApplicability of segmentation method is essentially required to control the newlinememory utilization effectively and to improve the execution speed The newlineutilization of memory is performed in an optimal manner by using a valid newlinerandom segmentation model which is based on event task represented in a newlineconsolidated way by its pattern cycle and repetition Execution is also made newlinefaster by using time related attributes other temporal databases Finally the newlineparameters time and memory complexity are analyzed to prove the newlinecorrectness of the proposed algorithm SSPMiner newline newline