arXiv subjects

Jingying Zeng

Publications and source records attributed to Jingying Zeng.

23 records · Page 2Linked to original sources

arXiv2024

Large Language Models for Social Networks: Applications, Challenges, and Solutions

Large Language Models (LLMs) are transforming the way people generate, explore, and engage with content. We study how we can develop LLM applications for online social networks. Despite LLMs' successes in other domains, it is challenging to develop LLM-based products for social networks for numerous reasons, and it has been relatively under-reported in the research community. We categorize LLM applications for social networks into three categories. First is knowledge tasks where users want to find new knowledge and information, such as search and question-answering. Second is entertainment tasks where users want to consume interesting content, such as getting entertaining notification content. Third is foundational tasks that need to be done to moderate and operate the social networks, such as content annotation and LLM monitoring. For each task, we share the challenges we found, solutions we developed, and lessons we learned. To the best of our knowledge, this is the first comprehensive paper about developing LLM applications for social networks.

cs.SI↗

arXiv2023

Let AI Entertain You: Increasing User Engagement with Generative AI and Rejection Sampling

While generative AI excels in content generation, it does not always increase user engagement. This can be attributed to two main factors. First, generative AI generates content without incorporating explicit or implicit feedback about user interactions. Even if the generated content seems to be more informative or well-written, it does not necessarily lead to an increase in user activities, such as clicks. Second, there is a concern with the quality of the content generative AI produces, which often lacks the distinctiveness and authenticity that human-created content possesses. These two factors can lead to content that fails to meet specific needs and preferences of users, ultimately reducing its potential to be engaging. This paper presents a generic framework of how to improve user engagement with generative AI by leveraging user feedback. Our solutions employ rejection sampling, a technique used in reinforcement learning, to boost engagement metrics. We leveraged the framework in the context of email notification subject lines generation for an online social network, and achieved significant engagement metric lift including +1% Session and +0.4% Weekly Active Users. We believe our work offers a universal framework that enhances user engagement with generative AI, particularly when standard generative AI reaches its limits in terms of enhancing content to be more captivating. To the best of our knowledge, this represents an early milestone in the industry's successful use of generative AI to enhance user engagement.

cs.HC↗

arXiv2022

A Survey of Causal Inference Frameworks

Causal inference is a science with multi-disciplinary evolution and applications. On the one hand, it measures effects of treatments in observational data based on experimental designs and rigorous statistical inference to draw causal statements. One of the most influential framework in quantifying causal effects is the potential outcomes framework. On the other hand, causal graphical models utilizes directed edges to represent causalities and encodes conditional independence relationships among variables in the graphs. A series of research has been done both in reading-off conditional independencies from graphs and in re-constructing causal structures. In recent years, the most state-of-art research in causal inference starts unifying the different causal inference frameworks together. This survey aims to provide a review of the past work on causal inference, focusing mainly on potential outcomes framework and causal graphical models. We hope that this survey will help accelerate the understanding of causal inference in different domains.

stat.ME↗

arXiv2022

Bayesian Causal Inference in Sequentially Randomized Experiments with Noncompliance

Scientific researchers utilize randomized experiments to draw casual statements. Most early studies as well as current work on experiments with sequential intervention decisions has been focusing on estimating the causal effects among sequential treatments, ignoring the non-compliance issues that experimental units might not be compliant with the treatment assignments that they were originally allocated. A series of methodologies have been developed to address the non-compliance issues in randomized experiments with time-fixed treatment. However, to our best knowledge, there is little literature studies on the non-compliance issues in sequential experiments settings. In this paper, we go beyond the traditional methods using per-protocol, as-treated, or intention-to-treat analysis and propose a latent mixture Bayesian framework to estimate the sample-average treatment effect in sequential experiment having non-compliance concerns.

stat.ME↗

arXiv2022

Semiparametric Estimation on Multi-treatment Causal Effects via Cross-Fitting

Causal inference is a critical research area with multi-disciplinary origins and applications, ranging from statistics, computer science, economics, psychology to public health. In many scientific research, randomized experiments provide a golden standard for estimation of causal effects for decades. However, in many situations, randomized experiments are not feasible in practice so that practitioners need to rely on empirical investigation for causal reasoning. Causal inference via observational data is a challenging task since the knowledge of the treatment assignment mechanism is missing, which typically requires non-testable assumptions to make the inference possible. For several years, great effort has been devoted to the research of causal inference for binary treatments. In practice, it is also common to use observational data on multiple treatment comparisons. Within the potential outcomes framework, we propose a generalized cross-fitting estimator (GCF), which generalizes the doubly robust estimator with cross-fitting for binary treatment to multiple treatment comparisons and provides rigorous proofs on its statistical properties. This estimator permits the use of more flexible machine learning methods to model the nuisance parts, and based on relatively weak assumptions, while there is still a theoretical guarantee for valid statistical inference. We show the asymptotic properties of the GCF estimators, and provide the asymptotic simultaneous confidence intervals that achieve the semiparametric efficiency bound for average treatment effect. The performance of the estimator is accessed through simulation study based on the common evaluation metrics generally considered in the causal inference literature.

stat.ME↗