基于金融微博情感与传播效果的股票价格预测

目前对微博情绪与金融预测之间关系的研究多数停留于诸如模式识别、语义分析、情感挖掘等文本挖掘技术,而较少研究微博情感传递过程。以金融微博文本情感挖掘和语义分析为基础,对相关的股票价格曲线进行拟合预测分析,包括对微博信息转播模型的研究和对微博情绪预测模型的研究。首先通过分析微博转播过程中的多个因素,包括转发情绪吸收、微博内容影响力、微博作者影响力、微博发布时间等,对模型自身进行拟合效果优化。其次,针对认证和非认证用户分类分析,并加入了转发次数的对其的再度影响,发现不同类型不同转发的用户对于股市曲线的影响滞后期不同。最后,在针对股市曲线变化的不同时期,对模型的拟合效果进行分析。给定金融市场某一特定关键词,采集了500,000多条金融微博及其相关用户信息。实验结果表明,新集成模型表现强于简单神经网络模型,而且是否为认证用户以及微博转发次数对微博滞后期的影响有所不同。此外,新模型的拟合效果,在股市上升期模型的拟合效果最好,下降期次之,平稳振荡期效果最差。

Weibo moods and propagation factors based stock prices prediction

At present, there are many studies on the relationship between Weibo sentiment and financial forecast. Most forecasting studies concern excessively on text mining techniques, such as pattern recognition, semantic or sentiment analysis, but neglect the procedure of moods dissemination. We provide an integrated framework, including the semantic mining, information transmission and propagating factors analysis, to predict stock prices more accurately. First, we select several factors in the dissemination process, such as emotional absorption of forwarding, influence of content and poster, release time, etc. to optimize the fitting effect of original model. Second, we classify users into two categories, verified or unverified users. And we also take the count of forwarding into account, checking its effect on stock prices fluctuation. Third, we compare the fitting effect of prediction models for different periods of the stock curve. Given a certain keyword related to financial market, we collected over 500,000 Micro-blogs and their user information from Weibo. Experiments demonstrate that our proposed integrated framework outperformed the simple neural network method. We observe that user category and the count of forwarding differ on the lag phase of influence. And more, we found that the model fitting effect were the best in the rising periods of stock prices curves, the second place in the declining and the worst in the fluctuating.