Litmaps Webinar with Faheem Ullah on Dec 6, 2023

Litmaps is positioned as a literature review “discovery and organization” tool that speeds up finding relevant papers while helping researchers keep collections current. The core workflow centers on building a citation-network map from a starting point, then expanding and monitoring that map as new literature emerges—reducing the manual work of searching, screening, and tracking sources.

The webinar frames literature reviews as a multi-stage process—defining questions, searching, filtering, extracting data, and analyzing—contrasting ad hoc reviews with systematic reviews that rely on explicit search strings, staged screening (title → abstract → full text), and structured extraction. Against that backdrop, Litmaps is described as most useful in the early, high-friction parts: locating papers and organizing them so later steps can proceed more efficiently. Rather than replacing human judgment, the tool is presented as a facilitator that improves coverage and reduces time spent on repetitive discovery tasks.

A key concept is citation-network-based recommendation. In “Seed Maps,” users start from one paper (a “seed”)—identified by DOI or other identifiers, or by searching for a topic plus “review”—and Litmaps generates a connected map of relevant articles. The visualization uses publication date along one axis and citation counts along another, letting users quickly spot older but highly cited “pivotal” work and newer papers with less citation history but more recent impact. The map’s connectedness is not limited to direct references and citations; it also considers highly connected papers in the broader network, aiming to balance relevance with diversity.

“Discover” expands beyond a single seed by taking multiple input papers and recommending additional literature using distinct algorithms. The webinar highlights three approaches: Top Connected (citation-network connectivity), Semantic Search (AI-driven similarity based on titles and abstracts), and Co-author Search (mapping relationships through shared authorship). Discover also includes practical controls such as filters for publication date ranges and keywords, plus the ability to save papers into “Collections” for ongoing organization by subtopic or author group.

To keep a review from going stale, “Monitor” automates recurring discovery. Users set alerts based on their existing papers and topics so new publications can be surfaced over time, addressing the common problem of missing recent work that can undermine a review’s completeness.

The session also adds concrete use cases from an academic perspective: using Litmaps to generate research directions and potential gaps, accelerating literature reviews through snowballing (up to two degrees of reference expansion), and helping identify a manageable set of “best” papers to start reading (often those that are both recent and highly cited). During Q&A, Litmaps’ underlying search is described as drawing from Semantic Scholar for paper retrieval, while recommendations in Seed/Discover rely on Litmaps’ own algorithms and citation-network or content similarity logic. The discussion emphasizes that results are deterministic in algorithm behavior, though coverage can change as databases update.

Overall, the webinar’s message is that Litmaps helps researchers build, expand, and maintain a structured evidence base—turning literature review discovery into a more navigable, less manual process—while still requiring researchers to validate relevance and extract meaning themselves.