Entity identification using deep learning models

摘要

In one embodiment, the method includes accessing in a social networking system a first set of entities that the user has interacted with and a second set of entities. A deep learning model is used to determine a first set of vector representations of the first set of entities. A target entity is selected from the first set of entities, and the vector representation of the target entity is deleted from the first set. The remaining vector representations in the first set are combined to determine the user's vector representation. A deep learning model is used to determine a second set of vector representations of the second set of entities. A similarity score between the user and each of the target entity and the entities in the second set of entities is calculated. A deep learning model is used to update the vector representation of entities in the second set of entities based on the similarity score.