TY - GEN
T1 - Developing machine learning models to automate news classification
AU - Singh, Roshan
AU - Chun, Soon Ae
AU - Atluri, Vijay
N1 - Funding Information:
This work partially supported by grants from NSF CNS 1747728, NSF CNS1624503, and NRF-Korea: 2017S1A3A2066084.
Funding Information:
This work was conducted during the author's sabbatical year at NYU Govlab from CUNY College of Staten Island.
Publisher Copyright:
© 2020 ACM.
PY - 2020/6/15
Y1 - 2020/6/15
N2 - Reading news articles is essential and critical for understanding the local, nation-wide, and global emerging and developing events, as well as understanding the citizens' demands and critics' opinions. However, with the explosion of social media as news channels, citizens and groups of professionals share news and opinions, which has been the territory of trained journalists, adding more news to process. News often comes with multimedia objects, and suffers from integrity issues, especially with the unreliable or false claims, so-called fake news or altered or alternative facts. These quantity, diversity, and integrity pose significant challenges in the information age, not only for the decision-makers, including policymakers, business leaders but also for individual citizens. This study focuses on how the machine learning classification algorithms could help the news classifications in different categories to easily access the needed category of news and to filter out the noisy and harmful news.
AB - Reading news articles is essential and critical for understanding the local, nation-wide, and global emerging and developing events, as well as understanding the citizens' demands and critics' opinions. However, with the explosion of social media as news channels, citizens and groups of professionals share news and opinions, which has been the territory of trained journalists, adding more news to process. News often comes with multimedia objects, and suffers from integrity issues, especially with the unreliable or false claims, so-called fake news or altered or alternative facts. These quantity, diversity, and integrity pose significant challenges in the information age, not only for the decision-makers, including policymakers, business leaders but also for individual citizens. This study focuses on how the machine learning classification algorithms could help the news classifications in different categories to easily access the needed category of news and to filter out the noisy and harmful news.
KW - BERT
KW - Big Data
KW - Deep Learning
KW - Machine Learning
KW - News classification
UR - http://www.scopus.com/inward/record.url?scp=85086890594&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85086890594&partnerID=8YFLogxK
U2 - 10.1145/3396956.3397001
DO - 10.1145/3396956.3397001
M3 - Conference contribution
AN - SCOPUS:85086890594
T3 - ACM International Conference Proceeding Series
SP - 354
EP - 355
BT - Proceedings of the 21st Annual International Conference on Digital Government Research
A2 - Eom, Seok-Jin
A2 - Lee, Jooho
PB - Association for Computing Machinery
T2 - 21st Annual International Conference on Digital Government Research: Intelligent Government in the Intelligent Information Society, DGO 2020
Y2 - 15 June 2020 through 19 June 2020
ER -