Ontology-based Financial Events Tracking

With the continuous development of computer technology, the Internet has become an indispensable information source of daily work and life. And because of the characteristics of network information itself, this information is mainly qualitative reference for users. Particularly in the financial field, the main target of the processing has always been the structured data. These quantitative data are very suitable for computer analysis. However, due to the characteristics of the financial field, the accurate analysis of unstructured information, especially the real-time tracking and detection of information implied in the important financial events, will also have a crucial impact on the structured data forecast. The research content mainly consists of the following aspects:1) Construction of financial ontology: The establishment of financial Ontology Library covering basic information of the stocks field.2) Ontology-based automatic annotation of network information and event extraction: Semantic annotation of the financial news web pages based on the ontology, indexing and extraction financial events of labled pages.3) Event temporal phrases extraction and standardization: Extract financial events'temporal phrases and normalized to the form that the computer can handle.Firstly, in this thesis, the financial news pages are analysed, the building methodology and the language to build financial ontology are summarized. Financial news repository is set up by building the financial news ontology. The terms in the repository is used to annotate the financial news webpage to achieve the indexing and extraction of financial news events. This thesis introduces the data structure and storage structure of financial news annotation, and expounds the data structures and algorithms used in annotation, indexing and extraction.After the establishment of the ontology annotation repository, this thesis further introduces the temporal phrases extraction and standardization with the ontology annotation repository, including temporal phrases'definition, classifications and other related concepts, describes the algorithms of the temporal phrases extraction and standardization.Finally, the evaluation standards of the financial news web page annotation and the temporal phrases extraction and standardization have been described in details. On the basis of a large number of experiments, the feasibility and effectiveness have been proved.