Reducing the Search Space in Literature-Based Discovery by Exploring Outlier Documents: a Case Study in Finding Links Between Gut Microbiome and Alzheimer’s Disease

Abstract

Literature-based discovery tools have been often used to overcome the problem of fragmentation of science and to assist researchers in their process of cross-domain knowledge discovery.In this paper we propose a methodology for cross-domain literature-based discovery that focuses on outlier documents to reduce the search space of potential cross-domain links and to improve search efficiency. In a previous study, literature mining tools OntoGen for document clustering and CrossBee for cross-domain bridging term exploration were combined to search for hidden relations in scientific papers from two different domains of interest, where the utility of the approach was demonstrated in a study involving PubMed papers about Alzheimer's disease and gut microbiome. This paper extends the approach by proposing a methodology, implemented as a repeatable workflow in a web-based text mining platform TextFlows, which enables easy access and execution of the methodology for the interested researcher.

Bisociative Literature-Based Discovery: Lessons Learned and New Word Embedding Approach

“…In terms of representation learning, our past LBD research that led to the lessons learned described in "Past LBD results and lessons learned" was based on using the standard TF-IDF weighted BoW vector representations of text documents [7,17,31,36]. On the other hand, the novel LBD methodology proposed in this paper in "Towards creative embeddings-based bisociative LBD" exploits contemporary representations of text documents using embeddings, given that current research in natural language processing demonstrates that representation learning using embeddings is much more effective than using the standard TF-IDF BoW vector representation.…”

“…These experimental results indicate that it is justified that the search for b-terms can be focused on outlier documents, which contain a large majority of b-terms. Consequently, by focusing the exploration on outlier documents, the effort needed for finding cross-domain links is substantially reduced, as it requires to explore a When applying OntoGen on the documents of the new application domain using the Alzheimer's disease-gut microbiome domain pair [7], the OntoGen method uses domains A and C, and builds a joint document set A ∪ C . With this intention, two individual sets of documents (e.g., titles, abstracts, or full texts of scientific articles), one for each domain under research (namely, literature A on Alzheimer's disease and literature C on gut microbiome), were automatically retrieved from the PubMed database.…”

“…This list included common terms and specific molecular factors and pathways, which were manually identified in titles, abstracts, and keywords from 42 papers obtained from PubMed search query (gut AND Alzheimer), 55 of which appeared also among the 4723 terms extracted by CrossBee. During the evaluation phase, the relevant papers for each b-term candidate were reviewed and searched for potential clues justifying further investigation, resulting from relevant b-term discoveries confirmed by the domain expert [7].…”

“…Compared to outlier document detection using OntoGen, an upgraded methodology proposed by Cestnik et al [7] was implemented in a reusable outlier-based LBD methodology in a web-based text-mining platform TextFlows 5 [29] that allowed us to construct and execute advanced textmining workflows. The workflow shown in Fig.…”

Highlights from the International Work-Conference on Bioinformatics and Biomedical Engineering (IWBBIO) – 2015 and 2016 Editions

“…For this reason, text mining and literature-based tools have become indispensable to assist researchers finding articles of interest. In this regard, a cross-domain knowledge discovery tool is described in this special issue by Cestnik et al [9]. This tool implements a search space reduction by eliminating potential outlier documents and, thus, improving search efficiency.…”