Text categorization models for identifying unproven cancer treatments on the web

Area of Science:

• Medical Informatics
• Computational Linguistics
• Public Health

Background:

• The internet facilitates the widespread dissemination of unproven and inaccurate medical claims.
• Patients with serious or untreatable conditions are particularly vulnerable to exploitation by false medical promises.
• Such misinformation can lead to severe financial, psychological harm, and delayed treatment with proven therapies.

Purpose of the Study:

• To explore the application of machine learning (ML) techniques for identifying web pages promoting unproven medical claims.
• To develop an automated method for detecting health misinformation online.

Main Methods:

• Utilized machine learning algorithms to analyze web content.
• Developed and evaluated models for the automatic identification of web pages containing unproven medical claims.

Main Results:

• The developed machine learning models demonstrated a high capability in automatically identifying web pages with unproven medical claims.
• Performance significantly surpassed existing web tools and state-of-the-art search engine technologies in detecting such content.

Conclusions:

• Machine learning offers a feasible and effective automated solution for detecting unproven medical claims on the internet.
• This approach can serve as a valuable tool to protect patients from health misinformation and exploitation.