As the capabilities of generative language models continue to advance, the implications of biases ingrained within these models have garnered increasing attention from researchers, practitioners, and the broader public. This article investigates the challenges and risks associated with biases in large-scale language models like ChatGPT. We discuss the origins of biases, stemming from, among others, the nature of training data, model specifications, algorithmic constraints, product design, and policy decisions. We explore the ethical concerns arising from the unintended consequences of biased model outputs. We further analyze the potential opportunities to mitigate biases, the inevitability of some biases, and the implications of deploying these models in various applications, such as virtual assistants, content generation, and chatbots. Finally, we review the current approaches to identify, quantify, and mitigate biases in language models, emphasizing the need for a multi-disciplinary, collaborative effort to develop more equitable, transparent, and responsible AI systems. This article aims to stimulate a thoughtful dialogue within the artificial intelligence community, encouraging researchers and developers to reflect on the role of biases in generative language models and the ongoing pursuit of ethical AI.
Detecting misinformation threads is crucial to guarantee a healthy environment on social media. We address the problem using the data set created during the COVID-19 pandemic. It contains cascades of tweets discussing information weakly labeled as reliable or unreliable, based on a previous evaluation of the information source. The models identifying unreliable threads usually rely on textual features. But reliability is not just what is said, but by whom and to whom. We additionally leverage on network information. Following the homophily principle, we hypothesize that users who interact are generally interested in similar topics and spreading similar kind of news, which in turn is generally reliable or not. We test several methods to learn representations of the social interactions within the cascades, combining them with deep neural language models in a Multi-Input (MI) framework. Keeping track of the sequence of the interactions during the time, we improve over previous state-of-the-art models.
The COVID-19 pandemic has gained worldwide attention and allowed fake news, such as ``COVID-19 is the flu,'' to spread quickly and widely on social media. Combating this coronavirus infodemic demands effective methods to detect fake news. To this end, we propose a method to infer news credibility from hashtags involved in news dissemination on social media, motivated by the tight connection between hashtags and news credibility observed in our empirical analyses. We first introduce a new graph that captures all (direct and \textit{indirect}) relationships among hashtags. Then, a language-independent semi-supervised algorithm is developed to predict fake news based on this constructed graph. This study first investigates the indirect relationship among hashtags; the proposed approach can be extended to any homogeneous graph to capture a comprehensive relationship among nodes. Language independence opens the proposed method to multilingual fake news detection. Experiments conducted on two real-world datasets demonstrate the effectiveness of our approach in identifying fake news, especially at an \textit{early} stage of propagation.
The detection of state-sponsored trolls acting in misinformation operations is an unsolved and critical challenge for the research community, with repercussions that go beyond the online realm. In this paper, we propose a novel approach for the detection of troll accounts, which consists of two steps. The first step aims at classifying trajectories of accounts' online activities as belonging to either a troll account or to an organic user account. In the second step, we exploit the classified trajectories to compute a metric, namely "troll score", which allows us to quantify the extent to which an account behaves like a troll. Experimental results show that our approach identifies accounts' trajectories with an AUC close to 99% and, accordingly, classify trolls and organic users with an AUC of 97%. Finally, we evaluate whether the proposed solution can be generalized to different contexts (e.g., discussions about Covid-19) and generic misbehaving users, showing promising results that will be further expanded in our future endeavors.
Human-centered AI considers human experiences with AI performance. While abundant research has been helping AI achieve superhuman performance either by fully automatic or weak supervision learning, fewer endeavors are experimenting with how AI can tailor to humans' preferred skill level given fine-grained input. In this work, we guide the curriculum reinforcement learning results towards a preferred performance level that is neither too hard nor too easy via learning from the human decision process. To achieve this, we developed a portable, interactive platform that enables the user to interact with agents online via manipulating the task difficulty, observing performance, and providing curriculum feedback. Our system is highly parallelizable, making it possible for a human to train large-scale reinforcement learning applications that require millions of samples without a server. The result demonstrates the effectiveness of an interactive curriculum for reinforcement learning involving human-in-the-loop. It shows reinforcement learning performance can successfully adjust in sync with the human desired difficulty level. We believe this research will open new doors for achieving flow and personalized adaptive difficulties.
Win prediction is crucial to understanding skill modeling, teamwork and matchmaking in esports. In this paper we propose GCN-WP, a semi-supervised win prediction model for esports based on graph convolutional networks. This model learns the structure of an esports league over the course of a season (1 year) and makes predictions on another similar league. This model integrates over 30 features about the match and players and employs graph convolution to classify games based on their neighborhood. Our model achieves state-of-the-art prediction accuracy when compared to machine learning or skill rating models for LoL. The framework is generalizable so it can easily be extended to other multiplayer online games.
Estimating the political leanings of social media users is a challenging and ever more pressing problem given the increase in social media consumption. We introduce Retweet-BERT, a simple and scalable model to estimate the political leanings of Twitter users. Retweet-BERT leverages the retweet network structure and the language used in users' profile descriptions. Our assumptions stem from patterns of networks and linguistics homophily among people who share similar ideologies. Retweet-BERT demonstrates competitive performance against other state-of-the-art baselines, achieving 96%-97% macro-F1 on two recent Twitter datasets (a COVID-19 dataset and a 2020 United States presidential elections dataset). We also perform manual validation to validate the performance of Retweet-BERT on users not in the training data. Finally, in a case study of COVID-19, we illustrate the presence of political echo chambers on Twitter and show that it exists primarily among right-leaning users. Our code is open-sourced and our data is publicly available.
To extract essential information from complex data, computer scientists have been developing machine learning models that learn low-dimensional representation mode. From such advances in machine learning research, not only computer scientists but also social scientists have benefited and advanced their research because human behavior or social phenomena lies in complex data. To document this emerging trend, we survey the recent studies that apply word embedding techniques to human behavior mining, building a taxonomy to illustrate the methods and procedures used in the surveyed papers and highlight the recent emerging trends applying word embedding models to non-textual human behavior data. This survey conducts a simple experiment to warn that common similarity measurements used in the literature could yield different results even if they return consistent results at an aggregate level.
With the recent development of technology, data on detailed human temporal behaviors has become available. Many methods have been proposed to mine those human dynamic behavior data and revealed valuable insights for research and businesses. However, most methods analyze only sequence of actions and do not study the inter-temporal information such as the time intervals between actions in a holistic manner. While actions and action time intervals are interdependent, it is challenging to integrate them because they have different natures: time and action. To overcome this challenge, we propose a unified method that analyzes user actions with intertemporal information (time interval). We simultaneously embed the user's action sequence and its time intervals to obtain a low-dimensional representation of the action along with intertemporal information. The paper demonstrates that the proposed method enables us to characterize user actions in terms of temporal context, using three real-world data sets. This paper demonstrates that explicit modeling of action sequences and inter-temporal user behavior information enable successful interpretable analysis.