Toxic comment classification has become an active research field with many recently proposed approaches. However, while these approaches address some of the task's challenges others still remain unsolved and directions for further research are needed. To this end, we compare different deep learning and shallow approaches on a new, large comment dataset and propose an ensemble that outperforms all individual models. Further, we validate our findings on a second dataset. The results of the ensemble enable us to perform an extensive error analysis, which reveals open challenges for state-of-the-art methods and directions towards pending future research. These challenges include missing paradigmatic context and inconsistent dataset labels.
University of Information Technology（信息技术大学）；Vietnam National University（越南国立大学）
We present our works on SemEval-2021 Task 5 about Toxic Spans Detection. This task aims to build a model for identifying toxic words in whole posts. We use the BiLSTM-CRF model combining with ToxicBERT Classification to train the detection model for identifying toxic words in posts. Our model achieves 62.23% by F1-score on the Toxic Spans Detection task.
Indian Institute of Technology, Kharagpur（印度理工学院卡拉格普尔分校）
The proliferation of online hate speech has necessitated the creation of algorithms which can detect toxicity. Most of the past research focuses on this detection as a classification task, but assigning an absolute toxicity label is often tricky. Hence, few of the past works transform the same task into a regression. This paper shows the comparative evaluation of different transformers and traditional machine learning models on a recently released toxicity severity measurement dataset by Jigsaw. We further demonstrate the issues with the model predictions using explainability analysis.
The presence of toxic content has become a major problem for many online communities. Moderators try to limit this problem by implementing more and more refined comment filters, but toxic users are constantly finding new ways to circumvent them. Our hypothesis is that while modifying toxic content and keywords to fool filters can be easy, hiding sentiment is harder. In this paper, we explore various aspects of sentiment detection and their correlation to toxicity, and use our results to implement a toxicity detection tool. We then test how adding the sentiment information helps detect toxicity…
The detection and identification of toxic comments are conducive to creating a civilized and harmonious Internet environment. In this experiment, we collected various data sets related to toxic comments. Because of the characteristics of comment data, we perform data cleaning and feature extraction operations on it from different angles to obtain different toxic comment training sets. In terms of model construction, we used the training set to train the models based on TFIDF and finetuned the Bert model separately. Finally, we encapsulated the code into software to score toxic comments in real…
In this paper we describe the methods we used for our submissions to the GermEval 2021 shared task on the identification of toxic, engaging, and fact-claiming comments. For all three subtasks we fine-tuned freely available transformer-based models from the Huggingface model hub. We evaluated the performance of various pre-trained models after fine-tuning on 80% of the training data with different hyperparameters and submitted predictions of the two best performing resulting models. We found that this approach worked best for subtask 3, for which we achieved an F1-score of 0.736.
University of Information Technology, Ho Chi Minh City, Vietnam（胡志明市信息技术大学）；Vietnam National University Ho Chi Minh City, Vietnam（越南国立大学胡志明市分校）
The rise of social media has led to the increasing of comments on online forums. However, there still exists invalid comments which are not informative for users. Moreover, those comments are also quite toxic and harmful to people. In this paper, we create a dataset for constructive and toxic speech detection, named UIT-ViCTSD (Vietnamese Constructive and Toxic Speech Detection dataset) with 10,000 human-annotated comments. For these tasks, we propose a system for constructive and toxic speech detection with the state-of-the-art transfer learning model in Vietnamese NLP as PhoBERT. With this s…
The challenge of automatic detection of toxic comments online has been the subject of a lot of research recently, but the focus has been mostly on detecting it in individual messages after they have been posted. Some authors have tried to predict if a conversation will derail into toxicity using the features of the first few messages. In this paper, we combine that approach with previous work on toxicity detection using sentiment information, and show how the sentiments expressed in the first messages of a conversation can help predict upcoming toxicity. Our results show that adding sentiment …
FH Aachen University of Applied Sciences（亚琛应用科学大学）
The availability of language representations learned by large pretrained neural network models (such as BERT and ELECTRA) has led to improvements in many downstream Natural Language Processing tasks in recent years. Pretrained models usually differ in pretraining objectives, architectures, and datasets they are trained on which can affect downstream performance. In this contribution, we fine-tuned German BERT and German ELECTRA models to identify toxic (subtask 1), engaging (subtask 2), and fact-claiming comments (subtask 3) in Facebook data provided by the GermEval 2021 competition. We create…
Toxic comment detection on social media has proven to be essential for content moderation. This paper compares a wide set of different models on a highly skewed multi-label hate speech dataset. We consider inference time and several metrics to measure performance and bias in our comparison. We show that all BERTs have similar performance regardless of the size, optimizations or language used to pre-train the models. RNNs are much faster at inference than any of the BERT. BiLSTM remains a good compromise between performance and inference time. RoBERTa with Focal Loss offers the best performance…
Toxic comment classification models are often found biased toward identity terms which are terms characterizing a specific group of people such as "Muslim" and "black". Such bias is commonly reflected in false-positive predictions, i.e. non-toxic comments with identity terms. In this work, we propose a novel approach to tackle such bias in toxic comment classification, leveraging the notion of subjectivity level of a comment and the presence of identity terms. We hypothesize that when a comment is made about a group of people that is characterized by an identity term, the likelihood of that co…
Toxic comments reduce the activity of volunteer editors on Wikipedia
RWTH Aachen University（亚琛工业大学）；University of Mannheim（曼海姆大学）；GESIS - Leibniz Institute for the Social Sciences（GESIS莱布尼茨社会科学研究所）；Complexity Science Hub Vienna（维也纳复杂科学中心）
The censorship of toxic comments is often left to the judgment of imperfect models. Perspective API, a creation of Google technology incubator Jigsaw, is perhaps the most widely used toxicity classifier in industry; the model is employed by several online communities including The New York Times to identify and filter out toxic comments with the goal of preserving online safety. Unfortunately, Google's model tends to unfairly assign higher toxicity scores to comments containing words referring to the identities of commonly targeted groups (e.g., "woman,'' "gay,'' etc.) because these identities…
While in real life everyone behaves themselves at least to some extent, it is much more difficult to expect people to behave themselves on the internet, because there are few checks or consequences for posting something toxic to others. Yet, for people on the other side, toxic texts often lead to serious psychological consequences. Detecting such toxic texts is challenging. In this paper, we attempt to build a toxicity detector using machine learning methods including CNN, Naive Bayes model, as well as LSTM. While there has been numerous groundwork laid by others, we aim to build models that p…
Toxic comments in online platforms are an unavoidable social issue under the cloak of anonymity. Hate speech detection has been actively done for languages such as English, German, or Italian, where manually labeled corpus has been released. In this work, we first present 9.4K manually labeled entertainment news comments for identifying Korean toxic speech, collected from a widely used online news platform in Korea. The comments are annotated regarding social bias and hate speech since both aspects are correlated. The inter-annotator agreement Krippendorff's alpha score is 0.492 and 0.496, res…
Athens University of Economic and Business（雅典经济与商业大学）；Google（谷歌）
Moderation is crucial to promoting healthy on-line discussions. Although several `toxicity' detection datasets and models have been published, most of them ignore the context of the posts, implicitly assuming that comments maybe judged independently. We investigate this assumption by focusing on two questions: (a) does context affect the human judgement, and (b) does conditioning on context improve performance of toxicity detection systems? We experiment with Wikipedia conversations, limiting the notion of context to the previous post in the thread and the discussion title. We find that contex…
Toxic content is one of the most critical issues for social media platforms today. India alone had 518 million social media users in 2020. In order to provide a good experience to content creators and their audience, it is crucial to flag toxic comments and the users who post that. But the big challenge is identifying toxicity in low resource Indic languages because of the presence of multiple representations of the same text. Moreover, the posts/comments on social media do not adhere to a particular format, grammar or sentence structure; this makes the task of abuse detection even more challe…
With surge in online platforms, there has been an upsurge in the user engagement on these platforms via comments and reactions. A large portion of such textual comments are abusive, rude and offensive to the audience. With machine learning systems in-place to check such comments coming onto platform, biases present in the training data gets passed onto the classifier leading to discrimination against a set of classes, religion and gender. In this work, we evaluate different classifiers and feature to estimate the bias in these classifiers along with their performance on downstream task of toxi…
Jalpaiguri Government Engineering College（杰尔拜古里政府工程学院）
Now-a-days, derogatory comments are often made by one another, not only in offline environment but also immensely in online environments like social networking websites and online communities. So, an Identification combined with Prevention System in all social networking websites and applications, including all the communities, existing in the digital world is a necessity. In such a system, the Identification Block should identify any negative online behaviour and should signal the Prevention Block to take action accordingly. This study aims to analyse any piece of text and detecting different…
Social network platforms are generally used to share positive, constructive, and insightful content. However, in recent times, people often get exposed to objectionable content like threat, identity attacks, hate speech, insults, obscene texts, offensive remarks or bullying. Existing work on toxic speech detection focuses on binary classification or on differentiating toxic speech among a small set of categories. This paper describes the system proposed by team Cisco for SemEval-2021 Task 5: Toxic Spans Detection, the first shared task focusing on detecting the spans in the text that attribute…
Federal University of São Carlos（圣卡洛斯联邦大学）；University of Sheffield（谢菲尔德大学）
Hate speech and toxic comments are a common concern of social media platform users. Although these comments are, fortunately, the minority in these platforms, they are still capable of causing harm. Therefore, identifying these comments is an important task for studying and preventing the proliferation of toxicity in social media. Previous work in automatically detecting toxic comments focus mainly in English, with very few work in languages like Brazilian Portuguese. In this paper, we propose a new large-scale dataset for Brazilian Portuguese with tweets annotated as either toxic or non-toxic…
Mapping Toxic Comments Across Demographics: A Dataset from German Public Broadcasting
Freie Universität Berlin（柏林自由大学）；Fraunhofer FOKUS（弗劳恩霍夫研究所（FOKUS））；MSB Medical School Berlin（柏林医学学校（MSB））；funk - Content-Netzwerk ARD / ZDF（ARD/ZDF内容网络）；InfAI Leipzig（莱比锡InfAI）
Rochester Institute of Technology（罗切斯特理工学院）；University of Wolverhampton（伍尔弗汉普顿大学）
This paper addresses the identification of toxic, engaging, and fact-claiming comments on social media. We used the dataset made available by the organizers of the GermEval-2021 shared task containing over 3,000 manually annotated Facebook comments in German. Considering the relatedness of the three tasks, we approached the problem using large pre-trained transformer models and multitask learning. Our results indicate that multitask learning achieves performance superior to the more common single task learning approach in all three tasks. We submit our best systems to GermEval-2021 under the t…
Classifiers tend to propagate biases present in the data on which they are trained. Hence, it is important to understand how the demographic identities of the annotators of comments affect the fairness of the resulting model. In this paper, we focus on the differences in the ways men and women annotate comments for toxicity, investigating how these differences result in models that amplify the opinions of male annotators. We find that the BERT model as-sociates toxic comments containing offensive words with male annotators, causing the model to predict 67.7% of toxic comments as having been an…
To identify and classify toxic online commentary, the modern tools of data science transform raw text into key features from which either thresholding or learning algorithms can make predictions for monitoring offensive conversations. We systematically evaluate 62 classifiers representing 19 major algorithmic families against features extracted from the Jigsaw dataset of Wikipedia comments. We compare the classifiers based on statistically significant differences in accuracy and relative execution time. Among these classifiers for identifying toxic comments, tree-based algorithms provide the m…
This paper tackles one of the greatest limitations in Machine Learning: Data Scarcity. Specifically, we explore whether high accuracy classifiers can be built from small datasets, utilizing a combination of data augmentation techniques and machine learning algorithms. In this paper, we experiment with Easy Data Augmentation (EDA) and Backtranslation, as well as with three popular learning algorithms, Logistic Regression, Support Vector Machine (SVM), and Bidirectional Long Short-Term Memory Network (Bi-LSTM). For our experimentation, we utilize the Wikipedia Toxic Comments dataset so that in t…
Facultat d’Informàtica de Barcelona (FIB)（巴塞罗那信息学院（FIB））；Universitat Politècnica de Catalunya (UPC - BarcelonaTECH)（加泰罗尼亚理工大学（UPC - 巴塞罗那理工））
This paper presents our submission to SemEval-2021 Task 5: Toxic Spans Detection. The purpose of this task is to detect the spans that make a text toxic, which is a complex labour for several reasons. Firstly, because of the intrinsic subjectivity of toxicity, and secondly, due to toxicity not always coming from single words like insults or offends, but sometimes from whole expressions formed by words that may not be toxic individually. Following this idea of focusing on both single words and multi-word expressions, we study the impact of using a multi-depth DistilBERT model, which uses embedd…
University of Information Technology（信息技术大学）；Vietnam National University Ho Chi Minh City（越南国立大学胡志明市分校）
The increment of toxic comments on online space is causing tremendous effects on other vulnerable users. For this reason, considerable efforts are made to deal with this, and SemEval-2021 Task 5: Toxic Spans Detection is one of those. This task asks competitors to extract spans that have toxicity from the given texts, and we have done several analyses to understand its structure before doing experiments. We solve this task by two approaches, Named Entity Recognition with spaCy library and Question-Answering with RoBERTa combining with ToxicBERT, and the former gains the highest F1-score of 66.…
Lack of moderation in online communities enables participants to incur in personal aggression, harassment or cyberbullying, issues that have been accentuated by extremist radicalisation in the contemporary post-truth politics scenario. This kind of hostility is usually expressed by means of toxic language, profanity or abusive statements. Recently Google has developed a machine-learning-based toxicity model in an attempt to assess the hostility of a comment; unfortunately, it has been suggested that said model can be deceived by adversarial attacks that manipulate the text sequence of the comm…
College of Information Science and Technology（信息科学与技术学院）；Pennsylvania State University（宾夕法尼亚州立大学）
Sociodemographic biases are a common problem for natural language processing, affecting the fairness and integrity of its applications. Within sentiment analysis, these biases may undermine sentiment predictions for texts that mention personal attributes that unbiased human readers would consider neutral. Such discrimination can have great consequences in the applications of sentiment analysis both in the public and private sectors. For example, incorrect inferences in applications like online abuse and opinion analysis in social media platforms can lead to unwanted ramifications, such as wron…
Télécom Paris（巴黎电信学院）；Institut Polytechnique de Paris（巴黎理工学院）；Stockholm University（斯德哥尔摩大学）；Google（谷歌）
Platforms that support online commentary, from social networks to news sites, are increasingly leveraging machine learning to assist their moderation efforts. But this process does not typically provide feedback to the author that would help them contribute according to the community guidelines. This is prohibitively time-consuming for human moderators to do, and computational approaches are still nascent. This work focuses on models that can help suggest rephrasings of toxic comments in a more civil manner. Inspired by recent progress in unpaired sequence-to-sequence tasks, a self-supervised …
Toxic conversations during software development interactions may have serious repercussions on a Free and Open Source Software (FOSS) development project. For example, victims of toxic conversations may become afraid to express themselves, therefore get demotivated, and may eventually leave the project. Automated filtering of toxic conversations may help a FOSS community to maintain healthy interactions among its members. However, off-the-shelf toxicity detectors perform poorly on Software Engineering (SE) datasets, such as one curated from code review comments. To encounter this challenge, we…
A large proportion of online comments present on public domains are constructive, however a significant proportion are toxic in nature. The comments contain lot of typos which increases the number of features manifold, making the ML model difficult to train. Considering the fact that the data scientists spend approximately 80% of their time in collecting, cleaning and organizing their data [1], we explored how much effort should we invest in the preprocessing (transformation) of raw comments before feeding it to the state-of-the-art classification models. With the help of four models on Jigsaw…
Vinod Gupta School of Management, IIT Kharagpur（印度理工学院卡拉格普尔分校维诺德·古普塔管理学院）
Toxic contents in online product review are a common phenomenon. A content is perceived to be toxic when it is rude, disrespectful, or unreasonable and make individuals leave the discussion. Machine learning algorithms helps the sell side community to identify such toxic patterns and eventually moderate such inputs. Yet, the extant literature provides fewer information about the sentiment of a prospective consumer on the perception of a product after being exposed to such toxic review content. In this study, we collect a balanced data set of review comments from 18 different players segregated…
Mediterranean Institute of Technology（地中海理工学院）；iCompass
The complete freedom of expression in social media has its costs especially in spreading harmful and abusive content that may induce people to act accordingly. Therefore, the need of detecting automatically such a content becomes an urgent task that will help and enhance the efficiency in limiting this toxic spread. Compared to other Arabic dialects which are mostly based on MSA, the Tunisian dialect is a combination of many other languages like MSA, Tamazight, Italian and French. Because of its rich language, dealing with NLP problems can be challenging due to the lack of large annotated data…
Deep learning methods are often difficult to apply in the legal domain due to the large amount of labeled data required by deep learning methods. A recent new trend in the deep learning community is the application of multi-task models that enable single deep neural networks to perform more than one task at the same time, for example classification and translation tasks. These powerful novel models are capable of transferring knowledge among different tasks or training sets and therefore could open up the legal domain for many deep learning applications. In this paper, we investigate the trans…
