Identifying Competitive Attributes Based on an Ensemble of Explainable Artificial Intelligence

Competitor analysis is a fundamental requirement in both strategic and operational management, the competitive attributes of reviewer comments are crucial determinant competitor approaches. Most studies have focused on identifying competitors or detecting comparative sentences, not attributes. Thus, authors propose method based explainable artificial intelligence (XAI) that can detect from consumers’ perspectives. They construct model to classify for each product calculate importance keyword during classification process. This assumption keywords significantly influence classification. The also an additional novel methodology combines various XAI techniques such as local interpretable model-agnostic explanations, Shapley additive logistic regression, gradient-based class activation map, layer-wise relevance propagation build robust calculating data sources.