A Dynamic Risk Assessment Approach Of Overseas Minerals Investment Project Based On Text Categorization

As the scale of overseas mining investment continuously increasing, it is more and more important to strengthen the study of risk assessment method of overseas mining investment project for supporting overseas mining investment decision-making. Risks of overseas mining investment projects have the characteristics of dynamic variation and concealment. And, the traditional risk assessment methods can’t dynamic reflect the changes of project risks that lead to the risk evaluation results are inaccurate. In order to realize the dynamic risk evaluation and improve the accuracy of risk assessment, based on the current situation of studies both at home and abroad, a dynamic risk assessment method was proposed in this study. It’s based on the improved text categorization algorithm, combining text categorization technology with traditional risk evaluation system. The method proposed by this study provides a new idea and way for evaluating the risk of overseas mining investment project.At present, the accuracy of text categorization is inadequate and existing feature selection algorithms of text categorization have some shortcomings. Firstly, this study focused on improving feature selection method to increase the accuracy of text categorization. Statistical analysis methods were used to obtain the formula of calculating how important the feature is to text categorization. The formula utilizes three factors of improved defined Frequency, Inter-class Concentration and Intra-class Dispersion to comprehensively evaluate the importance of the feature. By experiments on the Reuters-2158corpus and the Fudan University Chinese text classification corpus, the performance of proposed feature selection algorithm was compared with that of six traditional feature selection algorithms. And the results of the experiments showed that the method of this study can improve the classification accuracy of different classifier. Then, characteristics and risks of overseas mining investment project were deep studied. Based on the overseas mining investment’s actuality of our country, a dynamic risk assessment method based on text categorization for overseas mining investment projects was proposed, according to the characteristics of overseas mining investment project risk. In this method, text data which may be related to the project risks is processed by text categorization according to three dimensions of effective, relevance, credible degree. The results of text categorization processing are used to adjust the original risk assessment results obtained by traditional risk evaluation systems. As long as the adjusting process is repeated, dynamic evaluation of project risks is realized. Finally, taking one of overseas mining investment projects of China Nonferrous Metals Industry’s Foreign Engineering and Construction Co. as an example, the effectiveness of the dynamic risk assessment method was demonstrated.