New python package, ERTool, developed for efficient multi-source evidence fusion

Researchers from Peking University have developed ERTool, an open-source Python package designed to simplify the implementation of the Evidential Reasoning (ER) approach for multi-source evidence fusion. This tool addresses the challenges of integrating data from multiple sources in uncertain decision-making environments. The results are published in Health Data Science.

Multi-source evidence fusion plays a critical role in fields such as healthcare management, business analytics, and environmental risk assessment. However, the traditional application of the ER approach has been complicated, requiring expertise in coding. To overcome these challenges, Associate Research Professor Guilan Kong and her team at the National Institute of Health Data Science at Peking University designed ERTool, which automates the ER approach, making it accessible to a wider audience, including non-experts.

“Our goal was to make the ER approach more user-friendly, particularly for non-specialists,” explained Guilan Kong. “ERTool bridges the gap between complex algorithms and real-world applications, enabling researchers and professionals to integrate multi-source data for evidence-based decision-making more easily.”

The ERTool package simplifies the process of fusing evidence from different sources and addresses uncertainty in decision-making. It features a clean interface and high computational efficiency, making it a versatile tool for a range of applications. ERTool can be accessed via the Python Package Index or used through its online version, which supports real-time evidence fusion and result visualization.

In comparison with other systems like the Intelligent Decision System (IDS), ERTool is more accessible and easier to use, thanks to its open-source nature. It is freely available to the public, which enhances its potential for widespread use in various fields.

Moving forward, the research team plans to integrate a database management system (DBMS) into ERTool, which will allow it to handle larger volumes of evidence data. “Our ultimate goal is to make ERTool the leading solution for multi-source evidence fusion, continually evolving alongside the latest developments in evidential reasoning,” added Guilan Kong.