Construction of tax risk management evaluation system for listed enterprises from the perspective of big data

Currently, listed companies are facing problems in tax risk management, such as lagging tax policy updates, loopholes in internal control processes, and insufficient risk identification and warning capabilities. Big data technology can track policies in real time, dig deep into massive amounts of data to accurately identify risks, strengthen internal control processes, and help enterprises effectively control tax risks. Therefore, the study builds the tax risk management evaluation system of listed firms based on the big data technology backdrop in an attempt to raise the degree of tax risk management of enterprises. The experiment adopts the comprehensive assignment to calculate the index weights, and the evaluation object is comprehensively and deeply analyzed and evaluated through fuzzy mathematical methods. The study further adopts adaptive genetic algorithm and back-propagation neural network to construct a tax risk early warning method based on big data. The experiment applied the risk evaluation method proposed by the study to a listed company to obtain a tax risk value of 3.01, which existed a higher tax risk. With a higher risk value, the risk early warning model reached convergence at the 500th time with an error of 4%. Compared to the traditional back-propagation model, the improved model had lower error in the end. The correlation coefficient of the risk early warning model was 96% with high accuracy when the tax risk value was low. The proposed tax risk management evaluation and early warning methods for listed enterprises have strong practicality and application value, which can provide more accurate TAX risk management basis for enterprises and help them control tax risk effectively.