Application of Artificial Intelligence to Asset Pricing by Vietnamese Text Declaration

Asset pricing problem is an interesting problem by both academic and practical aspects. It has many practical applications such as pricing products on e-commerce transactions, pricing mortgage assets in credit activities,... In the era of data explosion, the application of big data analysis as well as artificial intelligence is an inevitable trend to produce more accurate predictive results. The article has applied artificial intelligence algorithms in asset pricing through text descriptions of the assets in Vietnamese. The proposed method uses Named Entity Recognition technique with a Recurrent Neural Network model in combination with Conditional Random Field model to extract asset features, thereby building a regression model to evaluate the price of assets based on the attribute set. The method works relatively well with a dataset of mobile phone descriptions with high accuracy.