Text Information Analysis Based Multiple Criteria Decision Model And Implementation

With rapid development of information technology, the Internet has entered the era of big data.Faced the wave of data explosion, we need a method which can analyze these massive data o supportour decision. And in real life, we usually take a lot of factors into consideration in decision making,which is a multiple criteria decision problem. Multiple criteria decision theory has a long history andis results are widely used in economics, management, medicine and many other fields.But throughoutrecent published research on multiple criteria decision, we can hardly find something about textinformation analysis. Therefore, in the context of big data, this paper attempts to combine textinformation analysis and multiple criteria decision theory to find out the knowledge and rules behindthe text to support customer’s decision. At the same time, do further analysis on customer’s decisionresults, calculating preference differences between customers and their decision rules for productrecommendation and opinions being feedback to the company who made the product.The main innovative points are as follows:1) Constructs a generic model which turns text data into numerical data, and defines how todetermine the semantic orientation, and proposes words’ clustering process;2) Innovatively based on the words’ clustering results generates the multi-objective decision-makingindex system, which has replaced traditional work of manual defining indexes; Besides, the computercan calculate the weight of the index automatically according to the words’ frequency, andfurthermore get the final weight after combining with policy-makers’ personal preference;3) Based on the results of multiple criteria decision, proposes preference analysis method, whichincludes the definition and quantifying of the preference difference and how to generate the graph;4) Calculates attribute reduction and decision rules by Rough Set Technology for further analysis ofthe law of the decision-makers’ behavior;5) Relying on the Jingdong Mall, chooses mobile phones as main experimental study to constructthe Multi-objective decision-making platform, with which customers can set their own index weightto choose products according to their preferences. Finally, we use excavated purchasing data toanalyze the case.