The news language materials happy pessimism Italy including controls the company with Taiwan gold to manage correlational of achievements appraisal target

碩士 === 銘傳大學 === 財務金融學系碩士在職專班 === 98 === This study applies the linguistic data mining technique to extract the information of financial events, financial management and administration related to the domestic financial holding companies from the public news in order to establish the quantified public...

Full description

Bibliographic Details
Main Authors: Chien-Hsun Chen, 陳建勳
Other Authors: Yang-Chang Lu
Format: Others
Language:zh-TW
Published: 2010
Online Access:http://ndltd.ncl.edu.tw/handle/81868644276632366854
Description
Summary:碩士 === 銘傳大學 === 財務金融學系碩士在職專班 === 98 === This study applies the linguistic data mining technique to extract the information of financial events, financial management and administration related to the domestic financial holding companies from the public news in order to establish the quantified public information indices. Through the analysis of these indices as well as all kinds of the indicators used by the financial holding companies for management performance evaluation, it is expected to predict the strong and weak points of the management performance indices of the financial holding companies according to the quantified news indices. By applying the corpus system, this study structures the news and builds up the optimistic and pessimistic emotional keywords from the news through effective document classification technique. Based on the optimistic and pessimistic keywords, this study further calculates the relevant quantified indices to conduct the correlation and regression analysis with the management performance indices used by the financial holding companies. The empirical result shows that the quantified indices MEDIA (degree of the public information disclosure), SRC (degree of net unweighted optimistic news) and SRW (degree of net weighted optimistic news) are all significantly positively correlated with some of the management performance indices respectively. Through correlation analysis, it is found that the MEDIA ((degree of the public information disclosure) is significantly negatively correlated with the cost ratio. However, the MEDIA (degree of the public information disclosure) will forecast the cost ratio of the financial holding company before the announcement of the quarterly report if the cost ratio is taken as the management performance evaluation index. The lower cost ratio will associate with higher MEDIA (degree of the public information disclosure). Moreover, the SRW (degree of net weighted optimistic news) and SRC (degree of net unweighted optimistic news) are significantly positively correlated with the return on equity, return on asset and earning per share. If the above three figures are taken as the management performance evaluation indices for the financial holding companies, the higher SRC (degree of net unweighted optimistic news) and SRW (degree of net weighted optimistic news) and SRC (degree of net unweighted optimistic news) will reflect higher return on equity, return on asset and earning per share on the quarterly financial report. The analysis further indicates that SRW index obtained after the weight added process of the optimistic and pessimistic keywords will have higher explanation capability comparing to SRC.