An Image-Based Hashtag Recommendation System as a Social Media Workflow Tool

Social Media is an obligatory choice to make for businesses in today's age to grow their brand. Attaining maximum reach on social media is pivotal for any business to gain consumers. This growth is directly proportional to the manner in which the business handles their social media. Hashtags are supposed to be extremely vital on social media platforms like Twitter, Instagram for easy retrieval, categorization and engagement of posts. But most businesses fail to make most of the benefits offered by usage of appropriate hashtags. Hence, in this research work we suggest a novel approach to recommend hashtags for image-based posts on social platforms using Deep Learning techniques. We also provide an in-depth methodology to build the essential dataset from scratch and develop the model for a real time web-based application. The proposed methodology has also been compared to a few other alternative approaches that we took. The experimental results showcase that our proposed approach is greatly significant in terms of providing hashtags for best reach as well as most suited to the post's image.