Duygu Analizi

Emotional analysis (sentiment analysis) is a natural language processing task, which aims to automatically identify certain emotions level, such as emotional tone that a text carries -usually positive, negative or neutral, sometimes more detailed angryness, joy or frustration. Simple approaches are prone to misinterpreting such methods as ironi, mecasis or binding sensitive expressions ("hiç was not conquered" when based on positive and negative word lists (externity-based methods). Instead of today’s systems, deep learning models consider the meaning of words within the context produce much more hit results, especially using Transformer-based classifiers.

Emotional analysis is an application that carries direct commercial value for businesses: companies can monitor customer reviews, product reviews, social media shares and survey results in large scale, and overall trends in brand perception can track the response given to a new product launch or a crisis instantly public sense in real time. This has become an indispensable tool for market research and brand reputation management, on the scale, which thousands of comments manually read. Applications associated with exchange movements of news and social media feeling analysis in the financial sector are becoming increasingly widespread.