Cognitive Assisted Blogger

Abstract: In this thesis a cognitive application is developed using IBMs cloud platform Bluemix and its Watson services. The goal of the application is to make it easier for bloggers and journalists to publish online during conferences. To reach the goal the application was designed with, among other things, functionalities for speech to text, publish on social media and automatically generated hashtags. The project was initially focused on developing a Bluemix hosted Node.js web application that would work on both iOS and Android devices. After constructing the Minimum Viable Product (MVP) for the web application the focus of the project changed and instead APIs were developed for the purpose to aid the development of native iOS and Android applications. The APIs were developed as Node.js applications with the help of API Connect and was afterwards hosted on Bluemix. This consisted of three separate APIs; the first API creates hashtags based on location and also tries to figure out who is speaking, to do this Facebook Graph and Alchemy was used. The second API makes it possible to save and get text from a cloud database and utilizes Cloudant. Lastly, the third API saves audio and also returns transcribed text and hashtags with the help of Object Storage, Watson Speech to Text and Alchemy.