deepSA2018 at SemEval-2018 Task 1: Multi-task Learning of Different Label for Affect in Tweets

Proceedings of the 12th International Workshop on Semantic Evaluation (SemEval-2018), pages 226–230 New Orleans, Louisiana, June 5–6, 2018. ©2018 Association for Computational Linguistics deepSA2018 at SemEval-2018 Task 1: Multi-task Learning of Different Label for Affect in Tweets 1Zi-Yuan Gao and 2Chia-Ping Chen Department of Computer Science and Engineering National Sun Yat-sen University Kaohsiung, Taiwan 1m053040030@student.nsysu.edu.tw 2cpchen@cse.nsysu.edu.tw Abstract This paper describes our system implementa- tion for subtask V-oc of SemEval-2018 Task 1: affect in tweets. We use multi-task learn- ing method to learn shared representation, then learn the features for each task. There are five classification models in the proposed multi- task learning approach. These classification models are trained sequentially to learn differ- ent features for different classification tasks. In addition to the data released for SemEval- 2018, we use datasets from previous SemEvals during system construction. Our Pearson cor- relation score is 0.638 on the official SemEval- 2018 Task 1 test set. 1 Introduction In recent years, people began to study how to cre- ate computational systems that process and under- stand the human languages. Today, people share their thoughts on social networks of the Internet, e.g. Facebook, Line, Twitter and so on. Thus, if the messages in the textual contents of social net- works can be extracted and summarized automat- ically via algorithms, it is possible to learn what people are interested in or are concerned with, and use such information to predict future market trends. Here we continue our previous works on the task 4 of SemEval-2017: Sentiment Analysis in Twitter (Rosenthal et al., 2017). SemEval-2017 subtask 4A is similar to task 1 of SemEval-2018: Affect in Tweets (Mohammad et al., 2018). They are challenging tasks as the messages on Twitter, called tweets, are short and informal. Further- more, in addition to noisy or incomplete texts, the emotional content of a tweet can be ambiguous and subjective. Affect in Tweets is an expanded version of WASSA-2017 shared task (Mohammad and Bravo-Marquez, 2017). The best system in WASSA-2017 is an ensemble of three sets of ap- proaches, including feed-forward neural network, multi-task deep learning and sequence modeling using CNNs and LSTMs (Goel et al., 2017). They attempt to use the idea of multi-task learning to ex- plore the notion of generalized or shared learning across different emotions. In this paper, we extend the idea with different label methods. The rest of this paper is organized as follows. In Section 2, we introduce our system. In Section 3, we describe the details of training and experimen- tal settings. In Section 4, we present the evaluation results along with our comments. 2 System Description 2.1 Baseline System Using RNN has become a very common tech- nique for various NLP tasks. There are many units for RNN-based model like simple RNN, gated recurrent units (GRU) (Chung et al., 2014), and long short-term memory (LSTM) (Hochreiter and Schmidhuber, 1997). For the baseline, we use LSTM as unit for its long-range dependency. Figure 1 shows the architecture of our baseline system. Our baseline system …

Content truncated to 500 words