Knowledge-Intensive Language Understanding for Explainable AI

AI systems have seen significant adoption in various domains. At the same time, further some domains is hindered by inability to fully trust an system that it will not harm a human. Besides, fairness, privacy, transparency, and explainability are vital developing systems. As stated Describing Trustworthy AI,aa.https://www.ibm.com/watson/trustworthy-ai. “Trust comes through understanding. How AI-led decisions made what determining factors were included crucial understand.” The subarea of explaining has come be known as XAI. Multiple aspects can explained; these include biases data might have, lack points particular region example space, fairness gathering data, feature importances, etc. However, besides these, critical human-centered explanations directly related decision-making, similar how domain expert makes based on “domain knowledge,” including well-established, peer-validated explicit guidelines. To understand validate system's outcomes (such classification, recommendations, predictions) lead system, necessary involve knowledge humans use. Contemporary XAI methods yet addressed enable decision-making expert. Figure 1 shows stages into real world.