اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدLOGAN: Local Group Bias Detection by Clustering
68
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Machine learning techniques have been widely used in natural language processing (NLP). However, as revealed by many recent studies, machine learning models often inherit and amplify the societal biases in data. Various metrics have been proposed to quantify biases in model predictions. In particular, several of them evaluate disparity in model performance between protected groups and advantaged groups in the test corpus. However, we argue that evaluating bias at the corpus level is not enough for understanding how biases are embedded in a model. In fact, a model with similar aggregated performance between different groups on the entire data may behave differently on instances in a local region. To analyze and detect such local bias, we propose LOGAN, a new bias detection technique based on clustering. Experiments on toxicity classification and object classification tasks show that LOGAN identifies bias in a local region and allows us to better analyze the biases in model predictions.
قيم البحث
اقرأ أيضاً
Leveraging unlabelled data through weak or distant supervision is a compelling approach to developing more effective text classification models. This paper proposes a simple but effective data augmentation method, which leverages the idea of pseudo-l
abelling to select samples from noisy distant supervision annotation datasets. The result shows that the proposed method improves the accuracy of biased news detection models.
Understanding a medical conversation between a patient and a physician poses a unique natural language understanding challenge since it combines elements of standard open ended conversation with very domain specific elements that require expertise an
d medical knowledge. Summarization of medical conversations is a particularly important aspect of medical conversation understanding since it addresses a very real need in medical practice: capturing the most important aspects of a medical encounter so that they can be used for medical decision making and subsequent follow ups. In this paper we present a novel approach to medical conversation summarization that leverages the unique and independent local structures created when gathering a patients medical history. Our approach is a variation of the pointer generator network where we introduce a penalty on the generator distribution, and we explicitly model negations. The model also captures important properties of medical conversations such as medical knowledge coming from standardized medical ontologies better than when those concepts are introduced explicitly. Through evaluation by doctors, we show that our approach is preferred on twice the number of summaries to the baseline pointer generator model and captures most or all of the information in 80% of the conversations making it a realistic alternative to costly manual summarization by medical experts.
The distance to the Large Magellanic Cloud (LMC) represents a key local rung of the extragalactic distance ladder. Yet, the galaxys distance modulus has long been an issue of contention, in particular in view of claims that most newly determined dist
ance moduli cluster tightly - and with a small spread - around the canonical distance modulus, (m-M)_0 = 18.50 mag. We compiled 233 separate LMC distance determinations published between 1990 and 2013. Our analysis of the individual distance moduli, as well as of their two-year means and standard deviations resulting from this largest data set of LMC distance moduli available to date, focuses specifically on Cepheid and RR Lyrae variable-star tracer populations, as well as on distance estimates based on features in the observational Hertzsprung-Russell diagram. We conclude that strong publication bias is unlikely to have been the main driver of the majority of published LMC distance moduli. However, for a given distance tracer, the body of publications leading to the tightly clustered distances is based on highly non-independent tracer samples and analysis methods, hence leading to significant correlations among the LMC distances reported in subsequent articles. Based on a careful, weighted combination, in a statistical sense, of the main stellar population tracers, we recommend that a slightly adjusted canonical distance modulus of (m-M)_0 = 18.49 +- 0.09 mag be used for all practical purposes that require a general distance scale without the need for accuracies of better than a few percent.
Dialogue systems play an increasingly important role in various aspects of our daily life. It is evident from recent research that dialogue systems trained on human conversation data are biased. In particular, they can produce responses that reflect
peoples gender prejudice. Many debiasing methods have been developed for various NLP tasks, such as word embedding. However, they are not directly applicable to dialogue systems because they are likely to force dialogue models to generate similar responses for different genders. This greatly degrades the diversity of the generated responses and immensely hurts the performance of the dialogue models. In this paper, we propose a novel adversarial learning framework Debiased-Chat to train dialogue models free from gender bias while keeping their performance. Extensive experiments on two real-world conversation datasets show that our framework significantly reduces gender bias in dialogue models while maintaining the response quality. The implementation of the proposed framework is released.
We study how masking and predicting tokens in an unsupervised fashion can give rise to linguistic structures and downstream performance gains. Recent theories have suggested that pretrained language models acquire useful inductive biases through mask
s that implicitly act as cloze reductions for downstream tasks. While appealing, we show that the success of the random masking strategy used in practice cannot be explained by such cloze-like masks alone. We construct cloze-like masks using task-specific lexicons for three different classification datasets and show that the majority of pretrained performance gains come from generic masks that are not associated with the lexicon. To explain the empirical success of these generic masks, we demonstrate a correspondence between the Masked Language Model (MLM) objective and existing methods for learning statistical dependencies in graphical models. Using this, we derive a method for extracting these learned statistical dependencies in MLMs and show that these dependencies encode useful inductive biases in the form of syntactic structures. In an unsupervised parsing evaluation, simply forming a minimum spanning tree on the implied statistical dependence structure outperforms a classic method for unsupervised parsing (58.74 vs. 55.91 UUAS).
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
