Jimin Wang

Xinpeng Qiu, Wang Yihu, Zhifeng Liu et al.

Scientific peer review generation has attracted increasing attention for reducing reviewing burdens and providing timely feedback. However, existing Large Language Model (LLM)-based methods often produce generic comments with insufficient evidence support and weak source traceability, while complex multi-agent systems incur high inference costs. To address these challenges, we propose EGTR-Review, an Evidence-Grounded and Traceable Review Generation framework via Multi-Agent Teacher Distillation. EGTR-Review first constructs a multi-agent teacher that performs structure-aware paper decomposition, key-element extraction, external scholarly evidence retrieval, evidence-state labeling, verification reasoning, and review synthesis. It then distills both intermediate reasoning trajectories and final review comments into a lightweight student model through task-prefix-driven multi-task learning. An evidence-weighted objective further reduces the influence of weak, missing, or non-verifiable supervision. Experiments on public peer-review datasets show that EGTR-Review (Student) outperforms strong prompt-based, fine-tuned, and structured/agentic baselines across automatic metrics, LLM-as-Judge evaluation, and human evaluation, while maintaining strong factual grounding and source traceability with substantially lower token consumption and inference time. Our code, prompts, configurations, and sample data are available on GitHub.

Liping Guo, Jimin Wang, Yanlong Zhao et al.

In the context of distributed fusion estimation, directly transmitting local estimates to the fusion center may cause a privacy leakage concerning exogenous inputs. Thus, it is crucial to protect exogenous inputs against full eavesdropping while achieving distributed fusion estimation. To address this issue, a noise injection strategy is provided by injecting mutually independent noises into the local estimates transmitted to the fusion center. To determine the covariance matrices of the injected noises, a constrained minimization problem is constructed by minimizing the sum of mean square errors of the local estimates while ensuring (ε, δ)-differential privacy. Suffering from the non-convexity of the minimization problem, an approach of relaxation is proposed, which efficiently solves the minimization problem without sacrificing differential privacy level. Then, a differentially private distributed fusion estimation algorithm based on the covariance intersection approach is developed. Further, by introducing a feedback mechanism, the fusion estimation accuracy is enhanced on the premise of the same (ε, δ)-differential privacy. Finally, an illustrative example is provided to demonstrate the effectiveness of the proposed algorithms, and the trade-off between differential privacy level and fusion estimation accuracy.

Jialong Chen, Jimin Wang, Ji-Feng Zhang

This paper proposes a differentially private gradient-tracking-based distributed stochastic optimization algorithm over directed graphs. In particular, privacy noises are incorporated into each agent's state and tracking variable to mitigate information leakage, after which the perturbed states and tracking variables are transmitted to neighbors. We design two novel schemes for the step-sizes and the sampling number within the algorithm. The sampling parameter-controlled subsampling method employed by both schemes enhances the differential privacy level, and ensures a finite cumulative privacy budget even over infinite iterations. The algorithm achieves both almost sure and mean square convergence for nonconvex objectives. Furthermore, when nonconvex objectives satisfy the Polyak-Lojasiewicz condition, Scheme (S1) achieves a polynomial mean square convergence rate, and Scheme (S2) achieves an exponential mean square convergence rate. The trade-off between privacy and convergence is presented. The effectiveness of the algorithm and its superior performance compared to existing works are illustrated through numerical examples of distributed training on the benchmark datasets "MNIST" and "CIFAR-10".

Jimin Wang, Yingjie Hu

A rich amount of geographic information exists in unstructured texts, such as Web pages, social media posts, housing advertisements, and historical archives. Geoparsers are useful tools that extract structured geographic information from unstructured texts, thereby enabling spatial analysis on textual data. While a number of geoparsers were developed, they were tested on different datasets using different metrics. Consequently, it is difficult to compare existing geoparsers or to compare a new geoparser with existing ones. In recent years, researchers created open and annotated corpora for testing geoparsers. While these corpora are extremely valuable, much effort is still needed for a researcher to prepare these datasets and deploy geoparsers for comparative experiments. This paper presents EUPEG: an Extensible and Unified Platform for Evaluating Geoparsers. EUPEG is an open source and Web based benchmarking platform which hosts a majority of open corpora, geoparsers, and performance metrics reported in the literature. It enables direct comparison of the hosted geoparsers, and a new geoparser can be connected to EUPEG and compared with other geoparsers. The main objective of EUPEG is to reduce the time and effort that researchers have to spend in preparing datasets and baselines, thereby increasing the efficiency and effectiveness of comparative experiments.

Yingjie Hu, Jimin Wang

Social media platforms, such as Twitter, have been increasingly used by people during natural disasters to share information and request for help. Hurricane Harvey was a category 4 hurricane that devastated Houston, Texas, USA in August 2017 and caused catastrophic flooding in the Houston metropolitan area. Hurricane Harvey also witnessed the widespread use of social media by the general public in response to this major disaster, and geographic locations are key information pieces described in many of the social media messages. A geoparsing system, or a geoparser, can be utilized to automatically extract and locate the described locations, which can help first responders reach the people in need. While a number of geoparsers have already been developed, it is unclear how effective they are in recognizing and geo-locating the locations described by people during natural disasters. To fill this gap, this work seeks to understand how people describe locations during a natural disaster by analyzing a sample of tweets posted during Hurricane Harvey. We then identify the limitations of existing geoparsers in processing these tweets, and discuss possible approaches to overcoming these limitations.

Jimin Wang, Yingjie Hu

Geoparsing is an important task in geographic information retrieval. A geoparsing system, known as a geoparser, takes some texts as the input and outputs the recognized place mentions and their location coordinates. In June 2019, a geoparsing competition, Toponym Resolution in Scientific Papers, was held as one of the SemEval 2019 tasks. The winning teams developed neural network based geoparsers that achieved outstanding performances (over 90% precision, recall, and F1 score for toponym recognition). This exciting result brings the question "are we there yet?", namely have we achieved high enough performances to possibly consider the problem of geoparsing as solved? One limitation of this competition is that the developed geoparsers were tested on only one dataset which has 45 research articles collected from the particular domain of Bio-medicine. It is known that the same geoparser can have very different performances on different datasets. Thus, this work performs a systematic evaluation of these state-of-the-art geoparsers using our recently developed benchmarking platform EUPEG that has eight annotated datasets, nine baseline geoparsers, and eight performance metrics. The evaluation result suggests that these new geoparsers indeed improve the performances of geoparsing on multiple datasets although some challenges remain.