Predict the Performance of Movie Box Office using Artificial Intelligence Technique
碩士 === 淡江大學 === 資訊工程學系碩士班 === 106 === Whether a movie can receive box office success involves many aspects of the problem, this study explores different group of input data to influence the result of predicting box office performance in movie industry. This study was conducted from Deep Learning met...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2018
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/an5d52 |
| Summary: | 碩士 === 淡江大學 === 資訊工程學系碩士班 === 106 === Whether a movie can receive box office success involves many aspects of the problem, this study explores different group of input data to influence the result of predicting box office performance in movie industry. This study was conducted from Deep Learning methods and conducted in-depth research on the box office prognostication technology.
The main source of the dataset is extracted from Opus Data database which covers the film release from 1997 to 2017. This study proposes an idea of labeling that the film will be grouped according to the production year firstly. The experimental shows the compresence of proposed labeling and related labeling and the result of different sets of input data. The results were evaluated in two terms of the accuracy: Bingo and 1-Away. The results show that adding specified input data can effectively improve the prediction results.
|
|---|