CS229 Project Report: Recommending movies and TV shows based on Facebook pro le data
نویسندگان
چکیده
The aim of this project is to learn how to make predictions on what genre of movie a user is likely to be interested in based on their raw Facebook data. We collaborated with other groups to collect a dataset of 10k unique users which was then parsed, stemmed and tokenized. We trained two predictors each using a di erent model, SVM and Naive Bayes. In order to correct for the skew in the training sets, we generated an independent unskewed dataset that was a subset of the original data, and compared performance of the classi ers using several metrics (accuracy, MCC, AUC). We then attempted several improvements, rst to the Naive Bayes model itself by using TF/IDF weighting and document vector length normalization, second to the features themselves by ltering tokens using Mutual Information. The combination of TF weighting and feature selection yields signi cantly better performance in terms of MCC and AUC.
منابع مشابه
CS229 Project Report: Automated photo tagging in Facebook
We examine the problem of automatically identifying and tagging users in photos on a social networking environment known as Facebook. The presented automatic facial tagging system is split into three subsystems: obtaining image data from Facebook, detecting faces in the images and recognizing the faces to match faces to individuals. Firstly, image data is extracted from Facebook by interfacing ...
متن کاملKeyword Based Recommender System for Electronic Products Using Weight Based Recommendation Algorithm Implemented on Hadoop
Recommender system is a topic which falls under the domain of information retrieval, data mining and machine learning. Recommender systems are widely used by famous websites like Amazon, Flipcart, Netflix, Facebook, twitter and many others. There are various types of recommender systems like collaborative filtering, content based filtering and hybrid recommender system. Recommender systems can ...
متن کاملQualitative Studying the Role of Cross-border TV Channel Movies on Housewives and Working Women Interact with Their Husbands
The present study was a qualitative study with the aim of studying the role of cross-border network drama films on how homemaking and working women interact with their husbands. The research society included working housewives living in Tehran through 2020, who watched movies on Persian-speaking channels based outside of the country. Among the aforementioned population, sampling was done in a t...
متن کاملTelevision Meets Facebook: The Correlation between TV Ratings and Social Media
This study examines the relationship between social media site Facebook and TV ratings drawing from audience factors of integration model of audience behavior. Based on context of Taiwan television network programs, this study collected measures for Facebook likes, shares, comments, posts for three genres of television shows and their Nielsen ratings over a period of eleven weeks, resulting in ...
متن کامل