Abstract
In the era of the explosion of digital content of large-scale self-media, user-friendly social platforms such as Twitter and Facebook, provide opportunities for people to express their ideas and opinions freely. Due to lack of restrictions, hateful speech and its exposure can have profound psychological impacts on society. Current social networking platform is over-reliant on the manual check, and it is labor-intensive and time-consuming. Although there are many machines learning methods for the detection of hate speech, short text with character limit on social platforms is more challenging for the detection of hate speech and offensive language. To address the problem of data sparsity, we have proposed a topic memory model for hate speech and offensive language detection (abbreviated as TM-HOL). Potential topics are generated with our encoder and decoder to enrich short text features. Two memory matrices correspond to the topic words and the text, and the hate feature matrix is used to learn the syntactic features. It is demonstrated that our proposed method is effective on three datasets, performing better weighted-F1.

Contributions
NTM-HOL: Potential topics are generated with our encoder and decoder to enrich short text features. The linear unit SELU layer in the model extracts topics. Neuron activations of NTM-HOL automatically converge towards zero mean and unit variance. This mechanism avoided the loss of essential information easily.
TMM-HOL: Two memory matrices and a hate feature matrix are proposed in TMM-HOL. Two memory matrices correspond to the topic words and the text, and the hate feature matrix in the modified computing layer is used to learn the syntactic features. It can make sentences and features merge better. This mechanism can solve the problem of losing the overall characteristics.

Code & Data
Code & Data: https://pan.baidu.com/s/1rCQCtB8t0O3An-SIfWLAYg Code: ccpe