TY - GEN
T1 - An effective news recommendation in social media based on users' preference
AU - Yuan, Xue
AU - Chen, Zhang
AU - Changzheng, Zhou
AU - Xun, Lin
AU - Qing, Li
PY - 2008/1/1
Y1 - 2008/1/1
N2 - In this paper, we have proposed a method to identify and track the drifted topics in the background of the social media through exploring the heated comments published, discussed, and voted by the participants of the social media. Based on this approach, we have further developed a way to optimize the recommendation of the relevant news to the readers of certain news by using the keywords generated from both news and comments. The challenge lies in how to select the keywords that are related with the drifted topics according to the users ' preference. In our work, we have utilized the number of votes received by a reader as an implicit feedback from the social media users in determining the quality of the comment. Then the keywords extracted from the comments are ranked based on both the quantity and the quality of the comments they appears in. Finally top-ranked keywords are selected and merged with the keywords representative of the original topics to retrieve the relevant news. Our experiment on news and comments from social media shows this approach is quite effective and promising.
AB - In this paper, we have proposed a method to identify and track the drifted topics in the background of the social media through exploring the heated comments published, discussed, and voted by the participants of the social media. Based on this approach, we have further developed a way to optimize the recommendation of the relevant news to the readers of certain news by using the keywords generated from both news and comments. The challenge lies in how to select the keywords that are related with the drifted topics according to the users ' preference. In our work, we have utilized the number of votes received by a reader as an implicit feedback from the social media users in determining the quality of the comment. Then the keywords extracted from the comments are ranked based on both the quantity and the quality of the comments they appears in. Finally top-ranked keywords are selected and merged with the keywords representative of the original topics to retrieve the relevant news. Our experiment on news and comments from social media shows this approach is quite effective and promising.
UR - http://www.scopus.com/inward/record.url?scp=70350646762&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70350646762&partnerID=8YFLogxK
U2 - 10.1109/ETTandGRS.2008.298
DO - 10.1109/ETTandGRS.2008.298
M3 - Conference contribution
SN - 9780769535630
T3 - 2008 International Workshop on Education Technology and Training and 2008 International Workshop on Geoscience and Remote Sensing, ETT and GRS 2008
SP - 627
EP - 631
BT - 2008 International Workshop on Education Technology and Training and 2008 International Workshop on Geoscience and Remote Sensing, ETT and GRS 2008
PB - IEEE Computer Society
T2 - 2008 International Workshop on Education Technology and Training and 2008 International Workshop on Geoscience and Remote Sensing, ETT and GRS 2008
Y2 - 21 December 2008 through 22 December 2008
ER -