Amobee at IEST 2018: Transfer Learning from Language Models
This work addresses emotion detection in social media for NLP applications, representing an incremental improvement in a specific shared task.
The paper tackled the problem of predicting implicit emotions in tweets with missing words, achieving first place in the WASSA 2018 shared task with a macro F1 score of 0.7145.
This paper describes the system developed at Amobee for the WASSA 2018 implicit emotions shared task (IEST). The goal of this task was to predict the emotion expressed by missing words in tweets without an explicit mention of those words. We developed an ensemble system consisting of language models together with LSTM-based networks containing a CNN attention mechanism. Our approach represents a novel use of language models (specifically trained on a large Twitter dataset) to predict and classify emotions. Our system reached 1st place with a macro $\text{F}_1$ score of 0.7145.