SYSTEM AND METHOD FOR RECOMMENDING STICKERS IN A SOCIAL NETWORKING APPLICATION

摘要

The present invention provides a method [200] and system [100] for recommending stickers in a social networking application. The system [100] includes a memory [102], at least one processor [104] connected to the memory [102], a clustering unit [106] and a sticker recommendation unit [108]. The clustering unit [106] is configured to receive at least one conversation history and extract at least one message set from the conversation history. The clustering unit [106] is further configured to train at least one model configured to compute a representation for a message in the message set and generate at least one message cluster based on the representation created by the trained model. The sticker recommendation unit [108] is configured to recommend one or more stickers based on a predicted cluster, wherein a classification model is used to predict the cluster.