Ethical Judgment Machine Learning Model Based On Text Analysis

Budget slack has always been a problem in the overall budget management of enterprises.This problem is one of the important reasons that make it difficult for budget management to effectively exert its control role.The behavior of budget slack is also an ethical issue in nature,and budget owners often face conflicts between personal interests and ethical norms.Whether to choose to create slack to maximize personal interests,or to choose to refuse slack to comply with universal ethics,there will be an ethical decision-making process.Therefore,this paper combines social media big data with machine learning methods in the context of the most classic budget slack problem in management accounting,and explores the construction of artificial intelligence algorithms that can identify and imitate human ethical judgment models.We first use crawler technology to obtain textual commentary data of Weibo social ethics events,and then improve the ethical judgment machine model by improving the SWNB text position analysis model algorithm proposed by Liu Kan et al.(2017).The artificial intelligence ethical judgment algorithm constructed by this method can accurately identify his ethical judgment model based on a person's comments on a typical ethical event,an d use this model to classify the sample.Then through the combination of experimental research with budget relaxation,explore the correlation between different ethical judgment models and budget slack.And the model can predict to some extent the budget slack that the budgeter may have when participating in the budget goal setting.The experimental results show that the budget participants under different ethical judgment models have significant differences in the ethical decision-making process of budget slack behavior and the budget slack in budget arrangement.Activists who are accustomed to applying the virtue ethics model and the fairness and justice model tend to have lower budget slack due to higher ethical sensitivity,and the actors of the rights and obligations model are in the middle.The actors of the utilitarian model are more concerned with personal interests than ethical sensitivities,and tend to have higher budget slack.The findings of this paper show that the ethical judgment model classification algorithm based on machine learning can be used to identify human ethical judgment models and can be used to predict people's behavior in ethical decision-making.This discovery also means that machines can be trained to imitate people for ethical judgment.Companies can also use ethical judgment machine learning models to give appropriate attention to employees' social media information,because many of the company's behavioral decisions(such as budget slack)are ethical activities.Enterprises can make a prior grasp through this information,thus providing reference for enterprises to implement relevant internal control management.