With increasing competition in the business then it takes a lot of data sources both internal and external to be processed and the results will produce the analysis required by these companies in making business decisions…

For internal data usually can be generated from the data warehouse system that already exists, but to external data derived from social media such as the data warehouse system can not directly produce them.
It is necessary for an application to perform an analysis of the brand and the specific keyword from social media sources such as Facebook, Twitter, Forum and others. The desired application must have the ability to pull data from social media, parsing and processing these data so that eventually formed a dashboard that displays the condition of our brand and keywords in the social media world.
In order to produce the results of the analysis regarding sentiment analytics and behavioral social media appropriately it is necessary also applications that can adapt to the language used, it is necessary for the application / module Natural Language Processing (NLP) that all forms of data conversations in social media can be interpreted with a more appropriate approach.
In its application, NLP is very dependent on the proper selection of keywords and also in defining trainer set made by analysts. For the manufacture of trainer set can not be done only one time but must be frequently tested and is updated within a certain time limit.