Research

Google Scholar ↗

Publications

  1. Feng Yang, Yuxin Hou, Xun Pang. 2025. Imputing International Trust Survey Data Using Large Language Models. Quarterly Journal of International Politics, 10(4), 1–31. (in Chinese) (online: 2026) (English title translated for this website.) DOI ↗
    Abstract

    English translation for this website.

    The rapid development of pretrained large language models has introduced generative modeling as an alternative approach to data processing in social science. This study examines missing-value imputation for international trust, a context-dependent latent subjective variable, using the longitudinal China Family Panel Studies (CFPS). It systematically compares large language models with established methods under different missing-data mechanisms and contrasts domestic and international models to assess the role of social context in training data and human feedback. Although generative approaches remain challenged by subjective survey-data imputation, they perform well under missing-not-at-random conditions, matching existing machine-learning methods and showing strong data fidelity. Domestic models outperform international models in understanding and simulating local political and social contexts. The findings identify opportunities for using large language models to address missing data in international-relations research and for improving the credibility and rigor of their use.

  2. Yuxin Hou, Junming Huang. 2025. Natural language processing for social science research: A comprehensive review. Chinese Journal of Sociology, 11(1), 121–157. DOI ↗ Abstract ↗
  3. Kun Yan, Yuxin Hou. 2023. Policy Diffusion of the Strategy of Reinvigorating China through Talent Development: Quantitative Research Based on Policy Literature from 2002 to 2021. Journal of National Academy of Education Administration, 7, 45–57, 77. (in Chinese)
    Abstract

    English translation for this website.

    Talent policy is a central instrument of China’s strategy of reinvigorating the country through talent development. Using bibliometric analysis of 2,163 policy documents issued by central and local governments from 2002 to 2021, this study constructs a policy-reference network and combines it with temporal keyword analysis to examine the intensity, breadth, speed and direction of policy diffusion. The issuing-agency reference network exhibits a three-level structure. Planning documents show greater diffusion intensity and breadth, with diffusion intensity continuing to grow throughout their validity periods; for opinions and decisions, growth lasts only four to seven years. Temporal keyword mapping identifies top-down, bottom-up, horizontal and follow-up diffusion patterns. The study characterizes the diffusion and development of talent policy and offers evidence for subsequent theoretical and practical work.

  4. Kun Yan, Yuxin Hou. 2022. The Construction of a Powerful Nation of Education in China: Policy Diffusion Process and Features. Tsinghua Journal of Education, 43(5), 79–92. (in Chinese)
    Abstract

    English translation for this website.

    This study uses bibliometric analysis to examine the diffusion of 12,356 policy documents on building China into a leading country in education, issued by central and local government departments between 2010 and 2021. It constructs policy-document and issuing-agency reference networks, analyzes intergovernmental relationships among policy issuers, and measures policy diffusion along four dimensions: intensity, breadth, speed and direction. By identifying patterns and characteristics of policy diffusion, the study provides a reference for future policy theory and practice.

  5. Kun Yan, Yuhui Jiang, Han Wu, Yuqi Zhang, Yuxin Hou. 2022. Promoting the Discipline-Major Dual Governance System Reform: Rational Analysis and Realization Path. Higher Education Review, 10(2), 242–256. (in Chinese)
    Abstract

    English translation for this website.

    Disciplines and degree programs coexist within universities. Their development supports research, social service and talent cultivation and is central to the qualitative development of higher education. This article distinguishes the meanings of disciplines and degree programs and explains their dialectical relationship. Drawing on their respective internal logics, it examines existing issues in the levels and modes of university education and proposes measures for reforming the dual governance system of disciplines and degree programs.

Working papers

  1. Yuxin Hou, Yuchen He, Chaojun Ni, Jia Yu, Junming Huang, Wenjun Mei. Comparing Opinion Structures Across Societies: Different in Surveys, Similar on Social Media. Working paper.
  2. Jia Yu, Rachel Ngai, Ting Chen, Yuxin Hou. Embedding Gender Norms. Working paper. (alphabetical order)
  3. Yuxin Hou, Airan Liu. Gendered Participation in Information Domains: Weibo Accounts’ Circulation and Expression of Public Affairs and Celebrity Culture. Working paper. (in Chinese) (English title translated for this website.)
  4. Junming Huang, Yu Xie, Yuxin Hou. Your friends are more polarized than you: scale-free network biases opinion perception and boosts online opinion polarization. Working paper.
  5. Yu Xie, Lemeng Liang, Yining Milly Yang, Junming Huang, Yueqi Xie, Yuxin Hou, Zheqing Ma, Weidong Wang, Yan Sun, Chunni Zhang. The Race between AI Superpowers: Social Factors Favoring China over the United States. Working paper.

AI assistant and conversation data

The guide is an AI character, not me. Opening interactions such as the die use authored public material without calling a model. The chat status indicates whether live conversation is available.

When live conversation is enabled, sending a question submits it and the current article identifier to this site's backend. The backend sends recent conversation, an index of earlier exchanges, any recalled records and retrieved public material to Alibaba Cloud Bailian to generate a response. Cloudflare Turnstile verifies visitors when a session is created; no account is required. Do not submit secrets or private information. The model and verification providers' data-processing policies apply.

The current browser tab stores the last ten interactions in session storage to support returning between reading and chat. Clearing the conversation removes those local records. Server sessions last one hour and retain up to twenty completed exchanges. The latest three are included directly in context; earlier records are retrieved when needed. Expired records are cleaned up on subsequent requests. “Start over” abandons the current session but does not immediately delete its server records. Across HTTPS pages in the same tab, session storage holds a one-hour session credential so the backend can restore recent context and an index of earlier exchanges. This is not the model provider’s API key. Expired credentials or unavailable storage require verification again; displayed history is not automatically imported into a new session.

The server also retains quota, budget and short-term request-limit records; connection identifiers are hashed with a secret key. The application does not keep additional question logs, but the session store contains the conversation text described above. Failed model requests produce operational records containing only a random request ID, fixed error category, processing stage, call count and whether partial output was shown; these contain no questions, answers, IP addresses or credentials. The AI's tools access public site material, not private files, personal chats or credentials. Citations do not guarantee that every sentence is correct: check the original sources and distinguish my views, quoted views and model inference.

Hi.

An AI character, not Yuxin herself. Opening interactions are available; live conversation is not connected yet.