CUFE at SemEval-2016 Task 4: A Gated Recurrent Model for Sentiment Classification

Mahmoud Nabil; Amir Atyia; Mohamed Aly

doi:10.18653/v1/S16-1005

CUFE at SemEval-2016 Task 4: A Gated Recurrent Model for Sentiment Classification

Mahmoud Nabil, Amir Atyia, Mohamed Aly

King Abdullah University of Science and Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

12 Scopus citations

Abstract

In this paper we describe a deep learning system that has been built for SemEval 2016 Task4 (Subtask A and B). In this work we trained a Gated Recurrent Unit (GRU) neural network model on top of two sets of word embeddings: (a) general word embeddings generated from unsupervised neural language model; and (b) task specific word embeddings generated from supervised neural language model that was trained to classify tweets into positive and negative categories. We also added a method for analyzing and splitting multi-words hashtags and appending them to the tweet body before feeding it to our model. Our models achieved 0.58 F1-measure for Subtask A (ranked 12/34) and 0.679 Recall for Subtask B (ranked 12/19).

Original language	English (US)
Title of host publication	Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016)
Publisher	Association for Computational Linguistics (ACL)
DOIs	https://doi.org/10.18653/v1/S16-1005
State	Published - Jul 14 2016

Access to Document

10.18653/v1/S16-1005

https://repository.kaust.edu.sa/bitstream/10754/618210/1/document.pdf

Cite this

@inproceedings{30e67a32c21b4f18b6a290d924fad221,

title = "CUFE at SemEval-2016 Task 4: A Gated Recurrent Model for Sentiment Classification",

abstract = "In this paper we describe a deep learning system that has been built for SemEval 2016 Task4 (Subtask A and B). In this work we trained a Gated Recurrent Unit (GRU) neural network model on top of two sets of word embeddings: (a) general word embeddings generated from unsupervised neural language model; and (b) task specific word embeddings generated from supervised neural language model that was trained to classify tweets into positive and negative categories. We also added a method for analyzing and splitting multi-words hashtags and appending them to the tweet body before feeding it to our model. Our models achieved 0.58 F1-measure for Subtask A (ranked 12/34) and 0.679 Recall for Subtask B (ranked 12/19).",

author = "Mahmoud Nabil and Amir Atyia and Mohamed Aly",

note = "KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This work has been funded by ITIDA{\textquoteright}s ITAC project number CFP65.",

year = "2016",

month = jul,

day = "14",

doi = "10.18653/v1/S16-1005",

language = "English (US)",

booktitle = "Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016)",

publisher = "Association for Computational Linguistics (ACL)",

address = "United States",

}

TY - GEN

T1 - CUFE at SemEval-2016 Task 4: A Gated Recurrent Model for Sentiment Classification

AU - Nabil, Mahmoud

AU - Atyia, Amir

AU - Aly, Mohamed

N1 - KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This work has been funded by ITIDA’s ITAC project number CFP65.

PY - 2016/7/14

Y1 - 2016/7/14

N2 - In this paper we describe a deep learning system that has been built for SemEval 2016 Task4 (Subtask A and B). In this work we trained a Gated Recurrent Unit (GRU) neural network model on top of two sets of word embeddings: (a) general word embeddings generated from unsupervised neural language model; and (b) task specific word embeddings generated from supervised neural language model that was trained to classify tweets into positive and negative categories. We also added a method for analyzing and splitting multi-words hashtags and appending them to the tweet body before feeding it to our model. Our models achieved 0.58 F1-measure for Subtask A (ranked 12/34) and 0.679 Recall for Subtask B (ranked 12/19).

AB - In this paper we describe a deep learning system that has been built for SemEval 2016 Task4 (Subtask A and B). In this work we trained a Gated Recurrent Unit (GRU) neural network model on top of two sets of word embeddings: (a) general word embeddings generated from unsupervised neural language model; and (b) task specific word embeddings generated from supervised neural language model that was trained to classify tweets into positive and negative categories. We also added a method for analyzing and splitting multi-words hashtags and appending them to the tweet body before feeding it to our model. Our models achieved 0.58 F1-measure for Subtask A (ranked 12/34) and 0.679 Recall for Subtask B (ranked 12/19).

UR - http://hdl.handle.net/10754/618210

UR - http://aclweb.org/anthology/S16-1005

U2 - 10.18653/v1/S16-1005

DO - 10.18653/v1/S16-1005

M3 - Conference contribution

BT - Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016)

PB - Association for Computational Linguistics (ACL)

ER -

CUFE at SemEval-2016 Task 4: A Gated Recurrent Model for Sentiment Classification

Abstract

Access to Document

Other files and links

Fingerprint

Cite this