A Movie Recommendation Method Based on Fractional Sentiment Analysis

Abstract

The invention discloses a movie recommendation method based on dimensional sentiment analysis. By crawling movie comment data, performing data preprocessing on the comment data, and extracting feature dimensions from the comment data after data preprocessing, the extracted Merge the feature dimensions, sort the merged feature dimensions, construct a sentiment dictionary suitable for the film field, use the constructed sentiment dictionary to perform sentiment analysis on movie comment data, obtain the movie genre model, and analyze the movie genre model Perform clustering operations to obtain recommended results and other steps. The advantages are: through the sentiment analysis of user movie reviews by feature dimension, the movie type model is calculated, which more accurately and comprehensively shows the characteristics of each feature dimension of the movie, thereby improving the quality level of the recommendation service, to a certain extent The problem of low accuracy of the traditional recommendation algorithm regardless of feature dimension has been improved.

Description

technical field [0001] The invention relates to the field of movie recommendation, in particular to a movie recommendation method based on dimensional sentiment analysis. Background technique [0002] With the rapid development of the Internet, different types of software and websites emerge in an endless stream. While it enriches our lives, it becomes more difficult to find the content that we are interested in from the vast data. Therefore, the recommendation system came into being. In the past, recommendation systems often used ratings or the overall sentiment of reviews to make recommendations, lacking in-depth mining of reviews. Reviews may cover multiple dimensions of information such as "actors", "directors", and "styles". Emotional tendencies are different. If recommendations are made only based on overall emotions, the accuracy rate will be low. In the past, movie recommendations mainly used the ratings given by users or the overall emotional tendency of movie revi...

AI Technical Summary

Problems solved by technology

Some producers may recruit some people to give high marks to their products for profit, resulting in unreliable recommendation results

Existing sentiment analysis can be roughly divided into two categories, dictionary-based methods and machine learning-based methods. Existing sentiment analysis algorithms have their own advantages and disadvantages, resulting in a single sentiment analysis method that cannot be applied to the movie recommendation system, reducing the system performance

Images