The tasks like scene identification and object classification interrelated to the concept detection. It involves scene
types as well as object categories which are relevant to the concepts. Visual concept has some negative aspects, while the
visual concept related to event analysis is carried out significant improvement. Visual concept is defined by human and it has
only one corresponding classifier in usual method are the negative aspects. This approach has proposed concept learning
algorithm for handling all these issues for social event detection in videos. System imparts a powerful automatic concept
mining algorithm with the help of N-gram internet services and flicker rather than defining visual concept manually. At the
same time depending on the learned visual concept, system repetitively finds out the multiple classifiers for each and every
concept. System gives a novel boosting concept learning algorithm, which increases the quality of being distinguishable.
Keywords — Classification, Event analysis, Video recognition, Visual concept detection.