基于天涯杂谈2015年全年帖子, 对其标题进行文本挖掘, 通过LDA主题 模型分类, 计算主题比率. 再通过对帖子的点击量, 回复量, 回复点击比, 持续热度各前100的帖 子进行词频统计, 得到上述4个指标的TOP100热帖. 进一步, 对比分析了 TOP100热帖的主题比率与全部帖子的主题比率. 文章的研究结 果可以捕捉到2015年天涯网友的热点关注方向, 结合情感分析技术, 研究 结果清晰地勾勒出天涯杂谈版块的网络舆情方向和网民态度.

In this paper, based on text mining and LDA topic model, we analyzed all posts on Tianya Zatan in 2015. We obtained TOP100 hot posts according to indexes such as clicking quantity, replying quantity, replies vs. clicks ratio and topics ratio. Furthermore, we compared the topics ratio of TOP100 and all posts. The empirical results can capture the focuses of Tianya Zatan netizens in 2015. Combined with emotional analysis technology, the research results clearly outline the public opinion direction and netizens' attitudes of Tianya Zatan.