Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications

作者： Dibakar Sigdel*1,2, Vincent Kyi*1,2, Aiden Zhang*1, Shaun P. Setty3, David A. Liem1,2,4, Yu Shi5, Xuan Wang5, Jiaming Shen5, Wei Wang1,6,7, JiaWei Han5, Peipei Ping1,2,4,6 1The NIH BD2K Center of Excellence in Biomedical Computing,University of California, Los Angeles, 2Department of Physiology,University of California, Los Angeles, 3Department of Pediatric and Adult Congenital Heart Surgery,Miller Children's and Women's Hospital and Long Beach Memorial Hospital, 4Department of Medicine/Cardiology,University of California, Los Angeles, 5NIH BD2K Program Centers of Excellence for Big Data Computing -- KnowEng Center, Department of Computer Science,University of Illinois at Urbana-Champaign (UIUC), 6Scalable Analytics Institute (ScAi),University of California, Los Angeles, 7Department of Computer Science,University of California, Los Angeles

简介：

Overview

This article presents a protocol for building a cloud-based phrase mining platform that facilitates the association of biomedical entities with specific diseases. The approach enhances efficiency and accessibility in biomedical research.

Key Study Components

Area of Science

• Biomedical literature analysis
• Text mining techniques
• Entity-category association

Background

• Manual evaluation of entity-category associations is time-consuming.
• Phrase mining tools can improve research efficiency.
• Cloud-based platforms enable broader access to text mining resources.
• Protocols can guide new users in implementing these tools.

Purpose of Study

• To automate the identification of phrase-category associations.
• To provide a systematic approach for analyzing biomedical literature.
• To enhance the usability of phrase mining tools for researchers.

Methods Used

• Step-by-step protocol for creating a text-cube from biomedical publications.
• Use of medical subject headings (MeSH) for defining categories.
• Implementation of Python scripts for data processing and analysis.
• Logging and debugging mechanisms to ensure process reliability.

Main Results

• Successful creation of a text-cube for document categorization.
• Automated mapping of entities to categories using MeSH descriptors.
• Generation of metadata and statistics for various age groups.
• Comparison of document counts across different subcategories.

Conclusions

• The protocol significantly streamlines the process of entity-category association.
• Cloud-based tools enhance accessibility for biomedical researchers.
• Future applications may include broader analyses across various biomedical domains.

Frequently Asked Questions