Document content extraction is a critical task in computer vision, underpinning the data needs of large language models (LLMs) and retrieval-augmented generation (RAG) systems. Despite recent progress, current document parsing methods have not been fairly and comprehensively evaluated due to the narrow coverage of document types and the simplified, unrealistic evaluation procedures in existing benchmarks. To address these gaps, we introduce OmniDocBench, a novel benchmark featuring high-quality annotations across nine document sources, including academic papers, textbooks, and more challenging cases such as handwritten notes and densely typeset newspapers. OmniDocBench supports flexible, multi-level evaluations—ranging from an end-to-end assessment to the task-specific and attribute-based analysis—using 19 layout categories and 15 attribute labels. We conduct a thorough evaluation of both pipeline-based methods and end-to-end vision-language models, revealing their strengths and weaknesses across different document types. OmniDocBench sets a new standard for the fair, diverse, and fine-grained evaluation in document parsing. Dataset and code are available at https://github.com/opendatalab/OmniDocBench.

We present a method for entity resolution that infers relationships between observed identities and uses those relationships to aid in mapping identities to underlying entities. We also introduce the idea of using graphlets for entity resolution. Graphlets are collections of small graphs that can be used to characterize the “role” of a node in a graph. The idea is that graphlets can provide a richer set of features to characterize identities. We validate our method on standard author datasets, and we further evaluate our method using data collected from Twitter. We find that inferred relationships and graphlets are useful for entity resolution.