کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
382182 | 660742 | 2015 | 8 صفحه PDF | دانلود رایگان |
![عکس صفحه اول مقاله: Study on SINA micro-blog personalized recommendation based on semantic network Study on SINA micro-blog personalized recommendation based on semantic network](/preview/png/382182.png)
• We divided the user interest more carefully.
• The tweets we recommended were from other uncorrelated users.
• The topic extracted by semantic network are more comprehensive and intuitive.
• We considered both tweet heat factor and tweeter authority factor in the recommendation model.
• The recommendation performance can be significantly improved.
In recent years, semantic network is applied to more and more research areas, such as Information Science areas. Differing from traditional users’ recommendations, the tweets’ recommendation in a micro-blog network has two crucial differences. One is high authority users or one’s special friends usually play a very active role in tweet-oriented recommendation. Micro-blog user will put the users his/her very interested into “special attention” group, and the topics discussed more in “special attention” group are more likely to be the user interested topic. The other is that users hope to obtain more relevant tweets about what he/she is interested in. Thus, this paper uses the k-cores analysis method to extract topics that users pay attention to, and employs the method of factor analysis to analyze index, and to extract the tweet heat factor and user authority factor. Besides, this paper intends to use the method of RS and linear regression to determine the parameters for balancing the value of the tweet heat factor and user authority factor. Finally, this paper manages to establish a timely personalized recommendation model based on semantic network for SINA tweets. According to the experimental results, the proposed method in this paper can effectively solve problems existing in micro-blog tweets in a personalized and timely recommendation way.
Journal: Expert Systems with Applications - Volume 42, Issue 10, 15 June 2015, Pages 4797–4804