there are few approaches I am aware of:

1) Salesforce Matching Rules

2) Salesforce Duplication Rules

3) Algorithms (Scoring, Matching key...)

4) For merging existing data - the best would be to select Salesforce AppExchange Duplicate Management App from Salesforce Partners

Our team prepared and extended this topic in a Blog: 

https://www.peeklogic.com/article/salesforce-duplicate-management/