Research On Related Entity Mining From Financial News

Financial market is a place where capital and production factors are arranged.It is the distribution center of a country's political,economic and cultural information.Financial markets attracts the attention of the media and a large number of people.The news media constantly publishing a large amount of news on the financial market every day.People pay close attention to the financial market through the media,and the news also affects the people's view on the financial assets.It is generally acknowledged that news have an impact on the financial markets.Mining useful information from mass news and positioning the impact of these news on financial assets is crucial to the operation of the financial markets and capital allocation.In the past,researchers studied the stock market based on the assumption that the stock entity is known to be affected by the news.However,due to this assumption,these methods inevitably ignore the news without stock entities,and many news without stock entities will also have a significant impact on financial markets.In order to solve this problem,this paper proposes a subgraph matching algorithm based on semantic paths.Matching subgraphs on a knowledge graph that collects a large amount of stock market information and matching the affected stock entities from the semantic level can make a comprehensive analysis on various news with or without entities.The main research work and achievements of this paper are as follows.The paper starts from structured data and using semi-structured and unstructured data as supplementary to construct knowledge graph on financial market.The knowledge graph contains most entities on financial market.This paper extract useful topics from financial news based on LDA topic analysis and construct a news graph.Then the paper proposed a method called subgraph matching algorithm based on semantic paths.Using the algorithm to find matching subgraph from kowledge graph and thus mining the related entities from financial news.Based on the matching result,we design experiments to compare our method with other algorithms to validate the accuracy of our method.And we also design an experiment to simulate truely investment and the algorithm get an excess return of 28.82% compare to the benchmark.The experiment validate the effectiveness of subgraph matching algorithm based on semantic paths and also validate the usefulness of our algorithm in investment.