Offensive Language Detection Using Multi-level Classification

Razavi, Amir H.; Inkpen, Diana; Uritsky, Sasha; Matwin, Stan

doi:10.1007/978-3-642-13059-5_5

Cited by 171 publications

(78 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Swearing is not necessarily impolite, inasmuch as offensive language is often used within the boundaries of what is considered situationally appropriate in discourse; further, some instances of swearing are neither polite nor impolite (Jay & Janschewitz, 2008). Offensive phrases could mocks or insult somebody or a group of people (attacks such as aggression against some culture, subgroup of the society, race or ideology in a tirade) (Rasavi, 2010). They will be offensive language if we use them for swearing or mocking other people.…”

Section: Introductionmentioning

confidence: 99%

Offensive Languages in Bad Boys 2

Mahayana¹

2017

KULTURISTIK

View full text Add to dashboard Cite

This research focuses on the forms of offensive language and the functions in social context found in Bad Boys 2 movie. The writer applied some methods, such as data source, data collection, and data analysis. The data source of this research was Bad Boys 2 movie's script. The movie is chosen because there are many expressions consisting of offensive language. The library research technique was used in collecting the data by quoting the sentences that support the topic of discussion. In analyzing the data, the writer applied the offensive language theory proposed by Timothy Jay (1992) in his book entitled Cursing in America and the theory of speech function proposed by Janet Holmes (1992) in his book entitled An Introduction to Sociolinguistics as the main theory. The analysis in this paper was also supported by other references which are related to this topic. The result of this research shows that there are only eight forms found in the data source. They are cursing, profanity, taboo, obscenity, vulgarity, slang, epithets, insult and slur.

show abstract

Section: Introductionmentioning

confidence: 99%

Offensive Languages in Bad Boys 2

Mahayana¹

2017

KULTURISTIK

View full text Add to dashboard Cite

show abstract

“…A query like "hore in bible" has a spelling mistake where hore refers to whore which makes the query inappropriate. Previous approaches [16,20,23,24] have focused on identifying offensive language or flames in the messages posted on online or social networking forums such as twitter and facebook. They mainly rely on the presence of strong offensive keywords or phrases and grammatical expressions.…”

Section: Query Completion Suggestions In Sesmentioning

confidence: 99%

“…It is pretty hard to extract topical features from search queries which are usually short and have less context. Razavi et al [16] detect flames (offensive/abusive rants) from text messages using a multi-level classification approach. They use a curated list of 2700 words, phrases and expressions denoting various degrees of flames and then used them as features for a two-staged Naive Bayes classifier.…”

Section: Related Workmentioning

confidence: 99%

Deep learning for detecting inappropriate content in text

Yenala

Jhanwar

Chinnakotla

et al. 2017

Int J Data Sci Anal

View full text Add to dashboard Cite

Today, there are a large number of online discussion fora on the internet which are meant for users to express, discuss and exchange their views and opinions on various topics. For example, news portals, blogs, social media channels such as youtube. typically allow users to express their views through comments. In such fora, it has been often observed that user conversations sometimes quickly derail and become inappropriate such as hurling abuses, passing rude and discourteous comments on individuals or certain groups/communities. Similarly, some virtual agents or bots have also been found to respond back to users with inappropriate messages. As a result, inappropriate messages or comments are turning into an online menace slowly degrading the effectiveness of user experiences. Hence, automatic detection and filtering of such inappropriate language has become an important problem for improving the quality of conversations with users as well as virtual agents. In this paper, we propose a novel deep learning-based technique for automatically identifying such inappropriate language. We especially focus on solving this problem in two application scenarios-(a) Query completion suggestions in search engines and (b) Users conversations in messengers. Detecting inappropriate language is challenging due to various natural language phenomenon such as spelling mistakes and variations, polysemy, contextual ambiguity and semantic variations. For identifying inappropriate query suggestions, we propose a novel deep learning architecture called "Convolutional Bi-Directional LSTM (C-BiLSTM)" which combines the strengths of both Convolution Neural Networks (CNN) and Bi-directional LSTMs (BLSTM). For filtering inappropriate conversations, we use LSTM and Bi-directional LSTM (BLSTM) sequential models. The proposed models do not rely on hand-crafted features, are trained end-end as a single model, and effectively capture both local features as well as their global semantics. Evaluating C-BiLSTM, LSTM and BLSTM models on real-world search queries and conversations reveals that they significantly outperform both pattern-based and other hand-crafted feature-based baselines.

show abstract

“…These paraphrases can be filtered out, when they are used in an application that prohibits such wording in the generated language. We do the filtering of the offensive expressions using a system from our previous work (Razavi et al 2010). …”

Section: Error Analysismentioning

confidence: 99%

A Bootstrapping Method for Extracting Paraphrases of Emotion Expressions From Texts

Keshtkar

Inkpen

2012

Computational Intelligence

Self Cite

View full text Add to dashboard Cite

Because paraphrasing is one of the crucial tasks in natural language understanding and generation, this paper introduces a novel technique to extract paraphrases for emotion terms, from nonparallel corpora. We present a bootstrapping technique for identifying paraphrases, starting with a small number of seeds. WordNet Affect emotion words are used as seeds. The bootstrapping approach learns extraction patterns for six classes of emotions. We use annotated blogs and other data sets as texts from which to extract paraphrases, based on the highest scoring extraction patterns. The results include lexical and morphosyntactic paraphrases, that we evaluate with human judges.

show abstract

Offensive Language Detection Using Multi-level Classification

Cited by 171 publications

References 21 publications

Offensive Languages in Bad Boys 2

Offensive Languages in Bad Boys 2

Deep learning for detecting inappropriate content in text

A Bootstrapping Method for Extracting Paraphrases of Emotion Expressions From Texts

Contact Info

Product

Resources

About