Malik Oyewale Salami — Metadata Quality & Research Integrity

Trust(Data) --> Model(Relationships) --> Understand(Science)

I am a doctoral candidate in Information Sciences at the School of Information Sciences, University of Illinois Urbana-Champaign (UIUC). Here is my iSchool page and CV.

My research sits at the intersection of science of science, data and information quality, scholarly communication, and computational social science. I develop computational methods for integrating, evaluating, and modeling large-scale scholarly data to understand how information infrastructures, technologies, institutions, and scholarly relationships shape the production, evaluation, and dissemination of knowledge. I promote open and reproducible research practices by making data, methods, and computational workflows transparent, reusable, and accessible.

I am advised by Professor Vetle I. Torvik of Torvik Research Group and co-advised by Professor Jodi Schneider of the InfoQuality Lab.

Research

Central question:

How can we build reliable scholarly data and computational models to better understand the structures, relationships, and practices that shape science?
I approach this question across interconnected research themes.

01 Data & Metadata Quality and Information Infrastructure ▾

Building trustworthy, open, and reusable foundations for studying science

I investigate the quality, consistency, and interoperability of scholarly data across information systems, while developing open and reusable datasets and computational workflows for studying science.

Retraction metadata Scholarly data Data integration Citation infrastructure Open science Reproducibility

02 Scholarly Network Modeling ▾

Understanding the social relationships underlying knowledge production

I develop computational models to study how scholarly relationships form and evolve within large-scale coauthorship networks.

Academic mentorship Academic genealogy Coauthorship networks Social computing Network analysis

03 Human & AI Reasoning over Scholarly Information ▾

Toward Transparent and Human-Aligned AI for Understanding Science

My emerging research examines how probabilistic models, large language models, and human judgment can complement one another when reasoning about scholarly relationships and metadata.Using academic mentorship as a testbed, I investigate whether LLMs can reason over contextual scholarly evidence to resolve ambiguous relationships that conventional models cannot confidently distinguish. I am particularly interested in comparing model inference, AI reasoning, and human judgment to understand where they agree, where they fail, and how they can be combined to create transparent and human-aligned scholarly information systems.

LLMs Explainable AI Scholarly information Human-in-the-loop systems