Author : Snehal R. Tanksale 1
Date of Publication :7th December 2016
Abstract: Sequential Pattern mining belongs to data mining techniques. Such techniques can applied to a sequential database to find out the correlation among list of ordered text documents. It is also used to identify time related user behavior in database. This can be useful in various areas such as, medical records, marketing etc. This is very challenging task to find out the sequential patterns from list of documents created by online users on social networking sites. The text mining is a popular solution to retrieve the information from text documents. Till now number of useful sequential patterns mining techniques from text documents has been developed and improvements are still going to improve the efficiency. Previous approaches do not consider the time stamp attribute of dataset. But it should be consider to improve the accuracy and efficiency of extracted information. So this paper makes survey of some recent approaches that focus on the sequential topic pattern mining for user abnormal behavior detection, also compare them on the basis of technique used and their advantages and disadvantages. It will be further used to improve the user behavior detection systems.
Reference :