Based
on the news of listed enterprises and the open public data, the Bd quantitative financial analysis system can make a
quantitative evaluation for listed enterprises from the dimensions of public
opinion and sentiment and can work out the attribution of public opinions based
on the contents of messages objectively.
The
algorithm is based on the data obtained from public channels and is uniform,
without any human intervention to the evaluation process and results. It is
aimed at being objective and fair to all evaluation
objects as much as possible.
*Note: Due to
integrity and promptness of data and accuracy of the algorithm, the evaluation
result might be inconsistent with the real conditions and is for reference
only. The evaluation result is not the view or
standpoint of Global Tone Communication Technology Co., Ltd. If
you have any objections to the evaluation results of the product, please
contact us via【support@gtcom-us.com】.
1. Sentiment algorithm:
According to the enterprise information obtained
from the public channels (including news and social media platforms), such as
product information, commercial cooperation and change of officer, the
automation algorithm is used for deeply processing the enterprise information,
and analyzing and calculating the same through multiple factors such as
objective sentiment factors, content of information, enterprise correlation and
attenuation time of the information. By virtual of combining all information
released by the news and the social channels for the whole same day and
combining the assigned media level weight, the final score is worked out.
*As the label is made by the sentiment algorithm
based on the corpus, the objective content will be mainly classified by
sentiments. After multiple manual rechecks are done and the extreme labels are
removed, we use the machine learning and training mode. However, the scoring
results might be impacted.