Loading...

Movie Success Prediction using Machine Learning Algorithms

Talei Farkhondeh, Mohammad Jafar | 2021

264 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54286 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Kianfar, Farhad
  7. Abstract:
  8. Briefly, the accomplishment of movies is based on their sales at the box office or gaining an achievement at the film festivals. Utilizing AI calculations and machine learning algorithms, various models have been proposed for the prediction of these achievements in the literature such as predicting the success of movies using regression models or the success of movies using discrete decision models is anticipated, and in another study the effect of social media data on sales investigated. The methods which are used to tackle these problems can be referred to as machine learning algorithms including neural networks, regression models, discrete decision-making models, public and experts opinions. In the regression models, an enormous number of boundaries such as actors, directors, previous records of the movie in film festivals, and their sales records are used. NLP algorithms can be utilized to detect patterns in words and captions used in social media to predict the success of movies. in this way, comments and articles about the movie are usually detected and extracted, and then using these algorithms sales or the success of the film at film festivals are predicted and analyzed. Another algorithm used in this field is the neural network algorithm that is developed using hidden layers to discover a connection between variables, which are mostly like regression models. The significance of this issue is for the movie producers; in this case, assuming they need a standard for estimating the success of a new film, these models will give them an awesome perspective. additionally, these models are used in media for attracting audiences subsequently gaining more profit, which makes predictions, particularly in the entertainment world, into an interesting subject. In this study, the data of Iranian cinema with the data of Iranian movie keyword trends (from google trends) are investigated to predict the interest in movies by models such as regression. Based on the results of this study, it can be highlighted the diminishing interest for Iranian movies, more interest in spring and summer seasons compared to cold seasons, and the direct connection between the demand of movies and interest over the time which the keywords are a trend. For modeling in this study, according to the collected data, regression models, and dimension reduction algorithms are used simultaneously and the outcomes are the result of utilizing these techniques
  9. Keywords:
  10. Machine Learning ; Cinema Economy ; Cinema Movies Sale ; Cinema ; Prediction Using Google Trends ; Cinema Movies Sale Prediction

 Digital Object List

 Bookmark

No TOC