國立高雄大學圖資館 |

語系: 繁體中文

說明(常見問題)

圖資館首頁

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Using machine learning techniques to...

Bora, Prachi.

Using machine learning techniques to predict the education level of Twitter users.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Using machine learning techniques to predict the education level of Twitter users.
作者:	Bora, Prachi.
面頁冊數:	67 p.
附註:	Source: Masters Abstracts International, Volume: 55-02.
附註:	Adviser: Tim Oates.
Contained By:	Masters Abstracts International55-02(E).
標題:	Computer science.
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1603712
ISBN:	9781339234298

Using machine learning techniques to predict the education level of Twitter users.
Bora, Prachi.

Using machine learning techniques to predict the education level of Twitter users. - 67 p.

Source: Masters Abstracts International, Volume: 55-02.

Thesis (M.S.)--University of Maryland, Baltimore County, 2015.

Over the past two decades, social media has been widely used to connect, socialize, and share content across the world. Platforms such as Twitter, Facebook, LinkedIn, and Google+ produce enormous amounts of data by and about users. This data can be used to predict hidden or latent attributes of a user like gender, age group, political inclination, regional origin, and many others, with applications in targeted advertising, recommendations, and personalization. While there is significant prior work on trying to predict attributes like age and gender, our work focuses on predicting the educational level of users from their Twitter feed. Specifically, we try to determine if a user is in high school or has (or is working on) a Ph.D. We use a machine learning approach to predict the educational level of Twitter users by looking at their tweet stream using Support Vector Machines, Perceptron, Logistic Regression, k-Nearest Neighbors, and Naive Bayes. We extract features from each class that clearly describe the characteristics of tweets from members of the classes. We also attempt to discover the abstract topics that occur in this dataset using Latent Dirichlet Allocation. These user attributes provide a very useful source of information for a variety of applications like advertising, search, friend recommendation, computational social science and many more.

ISBN: 9781339234298Subjects--Topical Terms:

199325
Computer science.

Using machine learning techniques to predict the education level of Twitter users.
LDR:02228nmm a2200265 4500 001 476040
005 20160418090144.5
008 160517s2015 ||||||||||||||||| ||eng d
020 $a 9781339234298
035 $a (MiAaPQ)AAI1603712
035 $a AAI1603712
040 $a MiAaPQ $c MiAaPQ
100 1 $a Bora, Prachi. $3 730283
245 1 0 $a Using machine learning techniques to predict the education level of Twitter users.
300 $a 67 p.
500 $a Source: Masters Abstracts International, Volume: 55-02.
500 $a Adviser: Tim Oates.
502 $a Thesis (M.S.)--University of Maryland, Baltimore County, 2015.
520 $a Over the past two decades, social media has been widely used to connect, socialize, and share content across the world. Platforms such as Twitter, Facebook, LinkedIn, and Google+ produce enormous amounts of data by and about users. This data can be used to predict hidden or latent attributes of a user like gender, age group, political inclination, regional origin, and many others, with applications in targeted advertising, recommendations, and personalization. While there is significant prior work on trying to predict attributes like age and gender, our work focuses on predicting the educational level of users from their Twitter feed. Specifically, we try to determine if a user is in high school or has (or is working on) a Ph.D. We use a machine learning approach to predict the educational level of Twitter users by looking at their tweet stream using Support Vector Machines, Perceptron, Logistic Regression, k-Nearest Neighbors, and Naive Bayes. We extract features from each class that clearly describe the characteristics of tweets from members of the classes. We also attempt to discover the abstract topics that occur in this dataset using Latent Dirichlet Allocation. These user attributes provide a very useful source of information for a variety of applications like advertising, search, friend recommendation, computational social science and many more.
590 $a School code: 0434.
650 4 $a Computer science. $3 199325
690 $a 0984
710 2 $a University of Maryland, Baltimore County. $b Computer Science. $3 730243
773 0 $t Masters Abstracts International $g 55-02(E).
790 $a 0434
791 $a M.S.
792 $a 2015
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1603712