Seasonal ARIMA Model Forecasting For Restaurant Reviews Word Frequency
碩士 === 國立虎尾科技大學 === 資訊工程系碩士班 === 107 === The aim of this research is to forecast word feature frequency from restaurant reviews by using time series analysis – Seasonal ARIMA model, which is one of the popular models in time series forecasting. In addition, Sentiment Analysis used for label each of...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2019
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/qd28ur |
| Summary: | 碩士 === 國立虎尾科技大學 === 資訊工程系碩士班 === 107 === The aim of this research is to forecast word feature frequency from restaurant reviews by using time series analysis – Seasonal ARIMA model, which is one of the popular models in time series forecasting. In addition, Sentiment Analysis used for label each of Positive, Negative, and Neutral review. After that, Term Frequency – Inverse Document Frequency or known as TF-IDF has been used to extract word feature and frequency for a positive and negative review. The result found ‘Great-Food’ is a feature of the word as a positive review and ‘Service-Food’ is a feature of negative reviews with the most frequency appeared and forecasting’s result show that it captured by word frequency seasonality.
|
|---|