Based on over a decade of deep data extraction and domain experience coupled with the latest artificial technology, Doctor Evidence has created a proprietary software in its DOC Search platform. DOC Search retrieves relevant medical documents from multiple sources and instantly performs dynamic analyses of the data contained in those sources.
Structured for ﬂexibility to incorporate multiple and federated data sources:
The machine learning architecture of the system dynamically links unstructured natural language, such as PubMed abstracts, with semi-structured information from other sources like clinical trial registrations and even news articles.
Powerful ontology system:
Enables inclusion of vast taxonomies that can manage synonyms and differences in jargon across the heterogeneous data sources.
Indexing of entities + the full pass of natural language processing on all documents:
Enables high-level analytics, such as co-occurrences and publication type detection.
State-of-the-art programming methodologies:
Functional programming, combined with industry-standard database technologies, provides the best of both worlds: a reliable foundation in an extremely ﬂexible environment.
Focus on interactivity:
Enables users to playfully explore the breadth and depth of a topic and provides a more enjoyable experience during a search session, while also performing as a highly functional tool for tasks such as protocol feasibility assessments and extracting search protocol elements.
Users can export a DOC Search Feasibility Report to obtain an overall summary of an evidence base and make the determination of whether there is sufficient data available to warrant a particular literature review.
Automatic identiﬁcation and linking together associated publications for the same trials (i.e., post-hocs, follow-up/extension studies, or subgroup analyses), including timelines.
Monitoring & signal detection:
Semi-automates the process of determining when a review should be updated. The “Signals” feature allows users to save a search and monitor for new literature of interest. Users can save an array of “Signals” and receive targeted alerts as soon as new articles are published, providing a real-time indication of whether a literature review should be updated.
Top thought leaders and/or inﬂuential authors (including collaboration networks) are identiﬁed by a powerful algorithm. Users can drill down on a speciﬁc author to retrieve that author’s contact information, affiliations, key interests, and top interventions reported in their publications.