System and method for analyzing result of clustering massive data

A task management apparatus for analyzing a result of clustering massive data, the task management apparatus comprising:

• a block generator configured to divide a clustered target file registered in a Hadoop Distribute File System (HDFS) into designated-sized blocks;
• an input split generator configured to combine the divided blocks, and generate an input split corresponding to a task pair for a reduce task for reducing input data;
• and
• an input split assigner configured to assign the generated input split to at least one distance calculation apparatus recognized in an identical network.

A distance calculation apparatus for analyzing a result of clustering massive data, the distance calculation apparatus comprising:

• a data acquirer configured to receive allocation of an input split corresponding to a task pair for a reduce task for reducing input data, and read all records of blocks included in the input split from a Hadoop Distribute File System (HDFS);
• a memory unit configured to store all the acquired records of the blocks;
• a calculator configured to calculate a distance sum for each of record between the blocks, and store the calculated distance sum for each record in the memory unit;
• and
• a data output unit configured to output the distance sum for each record.

A method of analyzing a result of clustering massive data, the method comprising:

• dividing a clustered target file into block of a pre-designated size;
• generating an input split corresponding to a task pair for a reduce task for reducing input data by combining the divided blocks;
• storing all records of block included in the input split into a memory, and outputting a distance sum for each of record;
• calculating a clustering significance verification index coefficient for each record by using the distance sum for each record;
• and
• defining a clustering significance verification index coefficient by averaging the clustering significance verification index coefficient for each record.