Emotion recognition has become an important area that plays a significant role in Human-Computer Interactions. Emotions can be expressed in many ways, such as facial expression and gestures, speech, and written text. A sufficient amount of work has been done regarding speech and facial emotion recognition, but a text-based emotion recognition system still needs researchers' attraction. With the increase in text information in social media posts, micro-blogs, news articles, etc., these contents can be beneficial to discover various aspects, including emotions. Emotion Detection in text documents is essentially a content-based classification problem involving concepts from the domains of Natural Language Processing and Machine Learning. In this paper, an overview of the emerging field of emotion detection from text is presented. The current generation of detection methods is usually divided into three main categories: keyword-based, learning-based, and hybrid recommendation approaches are discussed, along with their limitations. By examining the limitations of these approaches, the possible solution technique is suggested to improve emotion detection capabilities in practical systems, which emphasize human-computer interactions. To demonstrate the discussed process, a system is developed to accept file and text information and extract emotion from the text. This system is based on a neural network to classify and extract the text's emotion. This methodology can be beneficial in fields like emotion selling, social media analysis, integration with a chatbot for better customer interactions, etc.