Document Similarity algorithm


Homme avec son ordinateur sur du code

How do you define how similar documents are? Document Similarity algorithms identify documents that are semantically close and describe similar concepts.

Reading time: 2 minutes

Developing tools as a team.

Certipair asked for our help in setting up a Document Similarity algorithm. This is an innovative search engine technology. Our collaboration turned their algorithm into an intelligent solution.

Certipair offers a collaborative database. Its aim is to consolidate relations between doctors and patients. The organization has developed a platform guaranteeing better access to care. Patients are interested in seeing a secure, reassuring and efficient medical pathway. The aim of this database is to simplify health-related exchanges. Caregivers can easily search for and send short SMS or e-mail messages to their patients.

Améliorer le parcours de soin.

The development of this platform is based on two objectives. Thanks to its algorithm, the patient establishes a different type of relationship with his or her health care pathway. Personalized information (prevention, reminders, recommendations, etc.) is sent to the patient. The practitioner, for his part, saves time. This is made possible by our intelligent recommendation search engine module.

Working hand in hand with new technologies.

This Document Similarity algorithm is the result of recent scientific and technological advances. Indeed, we rely on NLP (Natural Language Processing), notably with Word2Vec and CamemBERT.

Other artificial intelligence modules are implemented to clean and organize this database in a completely autonomous way, using clustering algorithms.