A Review of Contract Entity Extraction

A contract is a binding document between two or more parties for executing any kind of activities that are defined clearly in its clauses. The parties who are assigned specific roles and rights must review and act accordingly until the contract expires. Automation of contract management is an emerging topic in various fields such as supply chain and legal domains. Recognition of various entities related to the document and their extraction are the main tasks to be performed for automating the contract management process. Named Entity Recognition is a well-known task in NLP that deals with the recognition of named entities such as person and date from a text. But traditional NER models perform poorly for domain-specific entity recognition and extraction such as legal and contract documents. For domain-specific entity recognition, we need to train the model with a dataset from the specific domain. In this paper, various approaches in the task of entity extraction from legal and contract documents are reviewed and discussed with the aim of proposing a new automated method for contract management using NLP.