Design and Implementation of Automatical Extension Method for

• 1、The State Key Laboratory of Networking and Switching Technology,Beijing University of Posts and Telecommunications,Beijing 100876

Abstract：In recent years, microblog has developed rapidly, and the number of microblog users is increasing. Behind the huge amount of users and data, there are huge commercial and social values. The emotional research and analysis of text information in the field of micro blog finance is to tap the potential business value of micro blog. Emotion dictionary for financial field is an important foundation to study financial field text. This research can be applied to many aspects of financial field, such as market analysis, user group research and so on. Based on so-pmi algorithm, this paper proposes the combination of LPA and PMI, designs and implements lpa-pmi algorithm, and uses micro-F1 as the evaluation tool to analyze the effectiveness of the algorithm in the expansion of emotion dictionary facing the financial field. The experimental results show that the financial dictionary constructed by the algorithm is more suitable for the emotional analysis in the financial field than the basic emotional dictionary, and the lpa-pmi algorithm is better than the so-pmi algorithm for the expansion of the emotional dictionary in the financial field. The validity of the lpa-pmi algorithm proposed in this paper is verified.

Keywords： Computer Technology Micro-Blog Sentiment Lexicon SO-PMI；LPA-PMI