A Survey on Hate Speech Detection using Natural Language Processing
Citations
728 citations
Cites background from "A Survey on Hate Speech Detection u..."
...In this survey [66], the authors provide a short, comprehensive, structured, and critical overview of the field of automatic hate speech detection in natural language processing....
[...]
611 citations
Cites background from "A Survey on Hate Speech Detection u..."
...A robust body of work has emerged trying to address the problem of hate speech and abusive language on social media (Schmidt and Wiegand, 2017)....
[...]
491 citations
Cites background or methods from "A Survey on Hate Speech Detection u..."
...Despite this large amount of work, it remains difficult to compare their performance [21], largely due to the use of different datasets by each work and the lack of comparative evaluations....
[...]
...State of the art primarily casts the problem as a supervised document classification task [21]....
[...]
...In addition, Knowledge-Based features such as messages mapped to stereotypical concepts in a knowledge base [8] and multimodal information such as image captions and pixel features [28] were used in cyber bully detection but only in very confined context [21]....
[...]
...It is widely recognised that a major limitation in this area of work is the lack of comparative evaluation [21]....
[...]
...[21] summarised several types of features used in the state of the art....
[...]
354 citations
351 citations
References
30,570 citations
"A Survey on Hate Speech Detection u..." refers background in this paper
...This work focuses on forecasting hit-and-run crimes from Twitter data by effectively employing semantic role labelling and event-based topic extraction (with LDA)....
[...]
...While Brown clustering produces hard clusters – that is, it assigns each individual word to one particular cluster – Latent Dirichlet Allocation (LDA) (Blei et al., 2003) produces for each word a topic distribution indicating to which degree a word belongs to each topic....
[...]
25,546 citations
20,077 citations
7,119 citations
"A Survey on Hate Speech Detection u..." refers background in this paper
...These paragraph embeddings (Le and Mikolov, 2014), which are internally based on word embeddings, have been shown to be much more effective than the averaging of word embeddings (Nobata et al....
[...]
3,336 citations
"A Survey on Hate Speech Detection u..." refers methods in this paper
...A standard algorithm for this is Brown clustering (Brown et al., 1992) which has been used as a feature in Warner...
[...]
...A standard algorithm for this is Brown clustering (Brown et al., 1992) which has been used as a feature in Warner and Hirschberg (2012)....
[...]