Paper Title
Identity Resolution for Crime Investigation Using AI Techniques

Abstract
Identity resolution is a concept of linking a particular identity based on its metadata to another identity in another database. Increasing efficiency in the field of Natural Language Processing (NLP), Text embedding techniques has become more robust which can help us find Semantic similarities between text information using these identity metadata. Therefore, in this paper, we have proposed a novel approach involving creating a vector or embedding based on the pre-trained model which provides a vector for identity-based text data based on pretraining knowledge. In addition to that, the distance calculation method is also used to find the distance between these text embeddings to find similarities between two identities. The proposed approach has been found to give good comparable performance to other existing techniques. Keywords - Semantic Similarity, BERT,ROBERTA, Identity Resolution, Cosine similarity