Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userFine-Grained Fairness Analysis of Abusive Language Detection Systems with CheckList
تحليل النزاهة الرخوة من أنظمة الكشف عن اللغة المسيئة مع قائمة مرجعية
417
0 0
0.0
(
0
)
Created by
Shamra Editor
Ask ChatGPT about the research
أظهرت أنظمة الكشف عن اللغة المسيئة الحالية التحيز غير المقصود تجاه ميزات حساسة مثل الجنسية أو الجنس. هذه قضية حاسمة، والتي قد تؤذي الأقليات والجماعات الممثلة تمثيلا ناقصا إذا تم دمج هذه الأنظمة في تطبيقات العالم الحقيقي. في هذه الورقة، نقوم بإنشاء اختبارات مخصصة من خلال أداة قائمة المراجعة (Ribeiro et al.، 2020) للكشف عن التحيزات داخل مصنفات اللغة المسيئة للغة الإنجليزية. نقارن سلوك نماذج استنادتين في بيرت، واحد مدرب على مجموعة بيانات الكلام الكراهية العامة والآخر في مجموعة بيانات للكشف عن الحرج. يوضح تقييمنا أنه على الرغم من أن المصنفات القائمة على بيرت تحقق مستويات عالية الدقة على مجموعة متنوعة من مهام معالجة اللغة الطبيعية، فإنها تؤدي بشكل سيء للغاية فيما يتعلق بالإنصاف والتحيز، لا سيما بشأن العينات التي تنطوي على الصور النمطية الضمنية، وتعبيرات عن الكراهية نحو الأقليات والسمات المحمية كما العرق أو الميل الجنسي. نطلق سراح كل من أجهزة الكمبيوتر المحمولة المنفذة لتوسيع اختبارات الإنصاف ومجموعات البيانات الاصطناعية التي يمكن استخدامها لتقييم تنظيم الأنظمة بشكل مستقل عن قائمة المراجعة.
Current abusive language detection systems have demonstrated unintended bias towards sensitive features such as nationality or gender. This is a crucial issue, which may harm minorities and underrepresented groups if such systems were integrated in real-world applications. In this paper, we create ad hoc tests through the CheckList tool (Ribeiro et al., 2020) to detect biases within abusive language classifiers for English. We compare the behaviour of two BERT-based models, one trained on a generic hate speech dataset and the other on a dataset for misogyny detection. Our evaluation shows that, although BERT-based classifiers achieve high accuracy levels on a variety of natural language processing tasks, they perform very poorly as regards fairness and bias, in particular on samples involving implicit stereotypes, expressions of hate towards minorities and protected attributes such as race or sexual orientation. We release both the notebooks implemented to extend the Fairness tests and the synthetic datasets usable to evaluate systems bias independently of CheckList.
References used
https://aclanthology.org/
rate research
Read More
The state-of-the-art abusive language detection models report great in-corpus performance, but underperform when evaluated on abusive comments that differ from the training scenario. As human annotation involves substantial time and effort, models th
at can adapt to newly collected comments can prove to be useful. In this paper, we investigate the effectiveness of several Unsupervised Domain Adaptation (UDA) approaches for the task of cross-corpora abusive language detection. In comparison, we adapt a variant of the BERT model, trained on large-scale abusive comments, using Masked Language Model (MLM) fine-tuning. Our evaluation shows that the UDA approaches result in sub-optimal performance, while the MLM fine-tuning does better in the cross-corpora setting. Detailed analysis reveals the limitations of the UDA approaches and emphasizes the need to build efficient adaptation methods for this task.
Since a lexicon-based approach is more elegant scientifically, explaining the solution components and being easier to generalize to other applications, this paper provides a new approach for offensive language and hate speech detection on social medi
a, which embodies a lexicon of implicit and explicit offensive and swearing expressions annotated with contextual information. Due to the severity of the social media abusive comments in Brazil, and the lack of research in Portuguese, Brazilian Portuguese is the language used to validate the models. Nevertheless, our method may be applied to any other language. The conducted experiments show the effectiveness of the proposed approach, outperforming the current baseline methods for the Portuguese language.
Abusive language detection has become an important tool for the cultivation of safe online platforms. We investigate the interaction of annotation quality and classifier performance. We use a new, fine-grained annotation scheme that allows us to dist
inguish between abusive language and colloquial uses of profanity that are not meant to harm. Our results show a tendency of crowd workers to overuse the abusive class, which creates an unrealistic class balance and affects classification accuracy. We also investigate different methods of distinguishing between explicit and implicit abuse and show lexicon-based approaches either over- or under-estimate the proportion of explicit abuse in data sets.
We introduce HateBERT, a re-trained BERT model for abusive language detection in English. The model was trained on RAL-E, a large-scale dataset of Reddit comments in English from communities banned for being offensive, abusive, or hateful that we hav
e curated and made available to the public. We present the results of a detailed comparison between a general pre-trained language model and the retrained version on three English datasets for offensive, abusive language and hate speech detection tasks. In all datasets, HateBERT outperforms the corresponding general BERT model. We also discuss a battery of experiments comparing the portability of the fine-tuned models across the datasets, suggesting that portability is affected by compatibility of the annotated phenomena.
As socially unacceptable language become pervasive in social media platforms, the need for automatic content moderation become more pressing. This contribution introduces the Dutch Abusive Language Corpus (DALC v1.0), a new dataset with tweets manual
ly an- notated for abusive language. The resource ad- dress a gap in language resources for Dutch and adopts a multi-layer annotation scheme modeling the explicitness and the target of the abusive messages. Baselines experiments on all annotation layers have been conducted, achieving a macro F1 score of 0.748 for binary classification of the explicitness layer and .489 for target classification.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
