The word sentiment has several meanings that can be synthesized into the Croatian verb noun opinion. Because we indirectly analyze the environment with our thoughts, writing reviews allows for a form of direct transmission of our own opinions. The review provides individuals with an insight into the quality of certain content, so it is extremely important in the entertainment industry. The most famous online site under the acronym IMDb offers interested individuals lists of movies, television series, domestic video productions and other components of the entertainment industry. Easily accessible lists are based solely on the opinion of the users of the site or their reviews. Therefore, they represent a key research basis for the problem of sentiment analysis. The most significant challenge in the field of natural language processing is the inability to recognize emotional tone, understand public opinion and recognize sarcasm, irony and the like. In this paper, I will analyze the English-Croatian translation versions of the collected sentiments from the IMDb page with the criterion of preserving the original sentiment. I will study the problems that arise when marking and translating sentiments from English to Croatian. Using non-commercial translation tools such as Google prevoditelj (eng. Google Translate) and Yandex, I will compare the quality of interlanguage translations. Finally, by analyzing the problems found, I will illustrate the challenges faced by the application of sentiment analysis, but also the cost-effectiveness of funding from the point of view of the entertainment industry.