Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
A New Method for Generating All Positive and Negative Association Rules(2011) ABSTRACT: Association Rule play very important role in recent scenario of data mining. But we have only generated positive rule, negative rule also useful in today data mining task. In this paper we are proposing “A new method for generating all positive and negative Association Rules” (NRGA).NRGA generates all association rules which are hidden when we have applied Apriori Algorithm. For representation of Negative Rules we are giving new name of this rules as like: CNR, ANR, and ACNR. In this paper we are also modify Correlation coefficient (CRC) equation, so all generate results are very promising. First we apply Apriori Algorithm for frequent itemset generation and that is also generate positive rules, after on frequent itemset we apply NRGA algorithm for all negative rules generation and optimize generated rules using Genetic Algorithm EXISTING SYSTEM: When data is saved in a distributed database, a distributed data mining algorithm is needed to mine association rules. Mining association rules in distributed environment is a distributed problem and must be performed using a distributed algorithm that doesn't need raw data exchange between participating sites. Distributed association rules mining (DARM), has been addressed by some researches and number of distributed algorithms have been proposed. Apriori is one of the most popular data mining approaches for finding frequent itemsets from transactional datasets. The Apriori algorithm is the main basis of many other well-known algorithms and implementations. The main challenge faced by the researchers in frequent itemset mining has been to reduce the execution time. Drawbacks: 1. The apriori takes more execution time for finding the frequent itemsets 2. The main drawback of the apriori algorithm is more candidate itemsets generation 3. The apriori is applicable only for offline transactional. PROPOSED SYSTEM: The point of this algorithm is that every site keeps a copy of trie locally, and they synchronize their data so that all local trie copies are the same at the end of each stage. After local support is counted, all sites share their support counts and determine the global support counts, in order to remove infrequent itemsets from their local trie. At the beginning, each site scans its local database independently, and determines the local count of items (1-itemsets). For this purpose, a vector is used to keep count of every item. Each site reads its local transaction records one by one and increase the count of items accordingly. At the end of this stage, sites synchronize their data to determine globally large 1-itemsets. Using L each site initializes its local trie copy; thus local trie copies are all alike at the end of the pass. Advantages: 1. The main advantage of trie based apriori is it takes less execution time than compare to apriori algorithm 2. Trie based apriori is applicable for online transactional datasets also 3. It reduces the database scans MODULES: Ø Association rule Ø Genetic algortihm Ø Optimization of association rule using GA SYSTEM REQUIREMENTS: Hardware requirements: · Processor · RAM · Hard Disk : Pentium IV : 512 MB : 40 GB Software requirements: · Operating System : Windows XP · Language used : Java · Back End : Oracle 10g