An analysis of gender bias studies in natural language processing

Addressing bias in artificial intelligence for public health surveillance

Components of artificial intelligence (AI) for analysing social big data, such as natural language processing (NLP) algorithms, have improved the timeliness and robustness of health data. NLP techniques have been implemented to analyse large volumes of text from social media platforms to gain insights on disease symptoms, understand barriers to care and predict disease outbreaks. However, AI-based decisions may contain biases that could misrepresent populations, skew results or lead to errors. Bias, within the scope of this paper, is described as the difference between the predictive values and true values within the modelling of an algorithm. Bias within algorithms may lead to inaccurate healthcare outcomes and exacerbate health disparities when results derived from these biased algorithms are applied to health interventions. Researchers who implement these algorithms must consider when and how bias may arise. This paper explores algorithmic biases as a result of data collection, labelling and modelling of NLP algorithms. Researchers have a role in ensuring that efforts towards combating bias are enforced, especially when drawing health conclusions derived from social media posts that are linguistically diverse. Through the implementation of open collaboration, auditing processes and the development of guidelines, researchers may be able to reduce bias and improve NLP algorithms that improve health surveillance.

Predicting drug-target binding affinity through molecule representation block based on multi-head attention and skip connection

Exiting computational models for drug-target binding affinity prediction have much room for improvement in prediction accuracy, robustness and generalization ability. Most deep learning models lack interpretability analysis and few studies provide application examples. Based on these observations, we presented a novel model named Molecule Representation Block-based Drug-Target binding Affinity prediction (MRBDTA). MRBDTA is composed of embedding and positional encoding, molecule representation block and interaction learning module. The advantages of MRBDTA are reflected in three aspects: (i) developing Trans block to extract molecule features through improving the encoder of transformer, (ii) introducing skip connection at encoder level in Trans block and (iii) enhancing the ability to capture interaction sites between proteins and drugs. The test results on two benchmark datasets manifest that MRBDTA achieves the best performance compared with 11 state-of-the-art models. Besides, through replacing Trans block with single Trans encoder and removing skip connection in Trans block, we verified that Trans block and skip connection could effectively improve the prediction accuracy and reliability of MRBDTA. Then, relying on multi-head attention mechanism, we performed interpretability analysis to illustrate that MRBDTA can correctly capture part of interaction sites between proteins and drugs. In case studies, we firstly employed MRBDTA to predict binding affinities between Food and Drug Administration-approved drugs and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication-related proteins. Secondly, we compared true binding affinities between 3C-like proteinase and 185 drugs with those predicted by MRBDTA. The final results of case studies reveal reliable performance of MRBDTA in drug design for SARS-CoV-2.

Gender Bias in Artificial Intelligence: Severity Prediction at an Early Stage of COVID-19

Artificial intelligence (AI) technologies have been applied in various medical domains to predict patient outcomes with high accuracy. As AI becomes more widely adopted, the problem of model bias is increasingly apparent. In this study, we investigate the model bias that can occur when training a model using datasets for only one particular gender and aim to present new insights into the bias issue. For the investigation, we considered an AI model that predicts severity at an early stage based on the medical records of coronavirus disease (COVID-19) patients. For 5,601 confirmed COVID-19 patients, we used 37 medical records, namely, basic patient information, physical index, initial examination findings, clinical findings, comorbidity diseases, and general blood test results at an early stage. To investigate the gender-based AI model bias, we trained and evaluated two separate models-one that was trained using only the male group, and the other using only the female group. When the model trained by the male-group data was applied to the female testing data, the overall accuracy decreased-sensitivity from 0.93 to 0.86, specificity from 0.92 to 0.86, accuracy from 0.92 to 0.86, balanced accuracy from 0.93 to 0.86, and area under the curve (AUC) from 0.97 to 0.94. Similarly, when the model trained by the female-group data was applied to the male testing data, once again, the overall accuracy decreased-sensitivity from 0.97 to 0.90, specificity from 0.96 to 0.91, accuracy from 0.96 to 0.91, balanced accuracy from 0.96 to 0.90, and AUC from 0.97 to 0.95. Furthermore, when we evaluated each gender-dependent model with the test data from the same gender used for training, the resultant accuracy was also lower than that from the unbiased model.

Metadata Normalization

Batch Normalization (BN) and its variants have delivered tremendous success in combating the covariate shift induced by the training step of deep learning methods. While these techniques normalize feature distributions by standardizing with batch statistics, they do not correct the influence on features from extraneous variables or multiple distributions. Such extra variables, referred to as metadata here, may create bias or confounding effects (e.g., race when classifying gender from face images). We introduce the Metadata Normalization (MDN) layer, a new batch-level operation which can be used end-to-end within the training framework, to correct the influence of metadata on feature distributions. MDN adopts a regression analysis technique traditionally used for preprocessing to remove (regress out) the metadata effects on model features during training. We utilize a metric based on distance correlation to quantify the distribution bias from the metadata and demonstrate that our method successfully removes metadata effects on four diverse settings: one synthetic, one 2D image, one video, and one 3D medical image dataset.

From Learning to Relearning: A Framework for Diminishing Bias in Social Robot Navigation

The exponentially increasing advances in robotics and machine learning are facilitating the transition of robots from being confined to controlled industrial spaces to performing novel everyday tasks in domestic and urban environments. In order to make the presence of robots safe as well as comfortable for humans, and to facilitate their acceptance in public environments, they are often equipped with social abilities for navigation and interaction. Socially compliant robot navigation is increasingly being learned from human observations or demonstrations. We argue that these techniques that typically aim to mimic human behavior do not guarantee fair behavior. As a consequence, social navigation models can replicate, promote, and amplify societal unfairness, such as discrimination and segregation. In this work, we investigate a framework for diminishing bias in social robot navigation models so that robots are equipped with the capability to plan as well as adapt their paths based on both physical and social demands. Our proposed framework consists of two components: learning which incorporates social context into the learning process to account for safety and comfort, and relearning to detect and correct potentially harmful outcomes before the onset. We provide both technological and societal analysis using three diverse case studies in different social scenarios of interaction. Moreover, we present ethical implications of deploying robots in social environments and propose potential solutions. Through this study, we highlight the importance and advocate for fairness in human-robot interactions in order to promote more equitable social relationships, roles, and dynamics and consequently positively influence our society.