Smart Dissemination by Using Natural Language Processing Technology

Background: Recently, WhatsApp has become the world's most popular text and voice messaging application with 1.5 billion users. A lot of WhatsApp Application Programming Interface (API) has also been established to be connected to other applications. On the other hand, the development of natural language processing (NLP) for WhatsApp messages has snowballed. There are extensive studies on the dissemination information using WhatsApp but the study on NLP involving data from WhatsApp is lacking. Objective: This study aims to implement NLP in smart dissemination applications by using WhatsApp API. Methods: We build a framework that embeds an intelligent system based on the NLP in WhatsApp API to disseminate a dynamic message. Some of the sentences are used to evaluate the accuracy of this application. Results: Smart dissemination consists of dynamic filter and dynamic content. Dynamic filter was conducted by using the POS tagger and clause statement. Meanwhile, dynamic content was built by using the replace MySQL function. There are twofold limitation: the application could not transform a message that matches rule with conjunction “dan”; has the same attribute before and after tag; and the maximum of the logical operator is one type for coordinating conjunction (AND/OR) in one sentence. Conclusion: Our framework can be used for dynamic dissemination of messages using dynamic message content and dynamic message recipient with an accuracy of 95% from twenty sample messages.