Gender Estimation on Social Media Using Recurrent Neural Network

With the development of instant messaging innovation and social media, protection has turned into a significant issue. There is a danger of one's record being hacked and utilized by the unknown person unconsciously. While doing texting on social media many people use abbreviations, short messages, emojis, images. We tried with different methods to gain the best accuracy in this research. In this paper, we will attempt to check the personality of the individual based on his/her composing style. We will explore the possibility of predicting the gender of a writer utilizing semantic proof. For this reason, term and style-based grouping strategies are assessed over an enormous accumulation of text messages. This study depicts the development of a huge, multilingual dataset named with gender, and examines factual models for deciding the gender of unknown Twitter clients. Twitter gives a basic method to clients to express sentiments, thoughts and assessments, makes the client produced content and related metadata, accessible to the network, and gives simple to utilize web and application programming interfaces to get to the information. The fundamental focal point of this paper is to gather the gender orientation of the client from unstructured data, including the username, screen name, depiction and picture, or by the client produced content