Artificial Intelligence-Driven Divorce Prediction: Integrating Psychometric, Financial, and Social Media Data

Mrs. Vidya V. Jagtap, Dr. Geetanjali Jindal

Divorce prediction, machine learning, artifi- cial intelligence, relationship viability, supervised learning, un- supervised learning, deep learning, psychometric assessment, algorithmic bias, data pr

The number of divorces around the world is still increasing, and people are very interested in how Artificial Intelligence (AI) and Machine Learning (ML) can be used to predict the outcomes of marriages. Divorce prediction systems use AI and ML to look at a lot of different things, like how people communicate, how they act emotionally, how much money they have, and how compatible they are with each other, to find patterns in relationships that are likely to end in divorce. By looking at these complicated details, predictive models can help couples who are having problems get help before it’s too late. This paper gives a thorough overview of AI/ML methods for predicting divorce, with a focus on supervised, unsupervised, and deep learning. These methods use information from a variety of sources such as psychometric tests, financial records, and even how people interact on social media, to predict how relationships will change over time. These predictive systems have effects that go beyond just one-on-one relationships. They could be useful in counselling, therapy, and the legal system. The paper also talks about the problems with data privacy, algorithmic biases, and model interpretability. It also looks at possible future directions that could make divorce prediction systems more accurate and ethical. In short, AI/ML-driven divorce prediction systems are a promising new way to help relationships stay healthy. They could change the way counselling is done and help keep society stable. Liu et al. Social media sites let people share their thoughts, feelings, and conversations in real time. This gives researchers a lot of data to work with, even though it’s not always clear, about marital satisfaction, relationship conflict, and possible signs of divorce. The researchers came up with semi-automated labelling methods and machine learning algorithms that can pick up on subtle signs of relationship distress, like sarcasm, coded language, and indirect references to problems in a marriage.