Where Do Your Citations Come From? Citation-Constellation: A Free, Open-Source, No-Code, and Auditable Tool for Citation Network Decomposition with Complementary BARON and HEROCON Scores

Mahbub Ul Alam · 2026

Read Where Do Your Citations Come From? Citation-Constellation: A Free, Open-Source, No-Code, and Auditable Tool for Citation Network Decomposition with Complementary BARON and HEROCON Scores by Mahbub Ul Alam (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-24216.txt.

Introduction

Alam introduces Citation-Constellation, a free no-code tool that splits a researcher's citations by network distance. Standard metrics count all citations alike. The BARON score counts only citations from outside the author's collaborative network. The HEROCON score gives graded credit to in-group citations by closeness. The gap between them flags dependence on an inner circle. The tool phases cover self-citations, co-authorship graphs, and affiliation matching. Alam stresses the scores are structural diagnostics, not quality measures, and should not drive hiring or funding choices.

Catalog record

CREATORMahbub Ul Alam
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE84008
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.24216

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Where Do Your Citations Come From? Citation-Constellation: A Free, Open-Source, No-Code, and Auditable Tool for Citation Network Decomposition with Complementary BARON and HEROCON Scores Thanks: The acronyms BARON and HEROCON serve as metaphors that encode the conceptual relationship between the scores, as detailed in Section 5 . Mahbub Ul Alam Abstract In this paper, I introduce Citation-Constellation, a fully operational and freely available tool 1 1 1 Citation-Constellation No-Code Tool: citation-constellation.serve.scilifelab.se for citation network analysis. Every claim presented here can be verified directly by the reader, with no installation, registration, or payment required. Users may enter an ORCID or OpenAlex ID to receive a complete citation network decomposition within minutes. The source code is publicly available 2 2 2 Citation-Constellation Source Code: github.com/citation-cosmograph/citation-constellation . Standard citation metrics treat all citations as equal, thereby obscuring the social and structural pathways through which scholarly influence actually propagates [ 82 , 35 ] . A citation from an independent researcher who discovers a paper through literature search carries a different epistemic weight than one from a direct co-author, a departmental colleague, or an editorial board member [ 42 , 78 ] . Yet conventional metrics such as the h-index, raw citation counts, and journal impact factors collapse these important distinctions into a single number. I introduce two complementary bibliometric scores that decompose a researcher’s citation profile according to network proximity between citing and cited authors. BARON (Boundary-Anchored Research Outreach Network score) provides a strict binary metric that counts only citations from outside the detected collaborative network, while HEROCON (Holistic Equilibrated Research Outreach CONstellation score) applies configurable graduated weights that assign partial credit to in-group citations based on relationship proximity. The gap between these two scores serves as a diagnostic indicator of inner-circle dependence. The accompanying open-source tool implements this decomposition through a phased detection architecture comprising (1) self-citation analysis, (2) co-authorship graph traversal, (3) temporal institutional affiliation matching via the Research Organization Registry (ROR), and (4) AI-agent-driven venue governance extraction using a locally deployed large language model 3 3 3 Citation-Astrolabe Source Code (venue governance database, under construction): citation-cosmograph/citation-astrolabe . The first three phases are fully operational and accessible via the provided web tool, while Phase 4 is currently under active implementation. The implementation places particular emphasis on ORCID-validated author identity resolution, an UNKNOWN classification for citations with insufficient metadata, and comprehensive audit trails that document every classification decision. The venue governance phase demonstrates how recent advances in locally deployable generative AI can automate bibliometric infrastructure tasks, such as editorial board extraction and entity resolution, that were previously intractable without dedicated research teams. A no-code web interface deployed on SciLifeLab Serve enables researchers to compute and visualize their scores without programming knowledge, installation, or registration. I present these scores as structural diagnostics that describe where in the social graph citations originate, explicitly not as quality indicators for research evaluation, and discuss their alignment with the responsible research assessment movement [ 21 , 35 ] . The graduated HEROCON weights remain experimental and require empirical calibration; I identify sensitivity analysis and cross-validation with citation motivation studies as priorities for future work. K eywords bibliometrics \cdot citation analysis \cdot citation network structure \cdot co-authorship network \cdot self-citation \cdot author disambiguation \cdot research impact measurement \cdot network proximity \cdot responsible research assessment \cdot open science \cdot research evaluation \cdot scientometrics \cdot scholarly communication \cdot venue governance \cdot institutional affiliation \cdot ORCID \cdot audit trail \cdot reproducibility \cdot large language model \cdot generative AI \cdot AI agent \cdot AI infrastructure \cdot automated information extraction \cdot natural language processing 1 Introduction The quantification of research impact represents a central concern of modern science policy, institutional evaluation, and individual career assessment. Metrics such as the h-index [ 36 ] , citation count, and journal impact factor have become deeply embedded in academic hiring, promotion, and funding decisions. Yet these metrics share a fundamental limitation: they treat all citations as equivalent signals of impact, regardless of the relationship between the citing and cited researchers. This limitation has been noted by numerous scholars [ 68 , 10 , 16 , 35 , 82 ] . A citation from an independent researcher who discovers a paper through literature search carries a different epistemic weight than a citation from a direct co-author, a colleague in the same department, or one that arrives through a venue where the cited researcher holds an editorial or governance role. The former represents genuine external reach, providing evidence that the work has found relevance beyond its originating community. The latter, while not illegitimate, reflects the natural amplification effects of collaborative networks, institutional proximity, and professional relationships. This phenomenon, the tendency for citation patterns to follow social network structure, has been extensively studied in the literature on invisible colleges [ 17 , 88 ] , co-authorship networks [ 52 , 53 , 29 , 51 ] , and citation homophily [ 78 , 77 ] . The core insight, that citation patterns are shaped by social network proximity, is well-established in network science and scientometrics. However, operationalizing this insight into a practical, auditable tool has been hindered by persistent challenges in author disambiguation, metadata quality, and the complexity of building comprehensive network detection layers that span self-citation, co-authorship, institutional affiliation, and venue governance. I address these challenges with an open-source tool that makes several key engineering contributions. First, I present Citation-Constellation, a phased detection architecture that progressively deepens network analysis from self-citation through co-authorship graphs, temporal affiliation matching, and venue governance detection, with each phase producing an independently meaningful and usable score. Second, the tool incorporates ORCID-validated author identity resolution to address the critical challenge of disambiguation error contaminating citation analysis. Third, an UNKNOWN classification honestly reports citations with insufficient metadata rather than silently defaulting them to EXTERNAL, preventing systematic bias against researchers with poor metadata coverage. Fourth, comprehensive audit transparency ensures that every classification decision is documented with a human-readable rationale in a structured JSON file, enabling full reproducibility and contestability. Fifth, the Phase 4 pipeline employs an AI-agent-driven venue governance extraction system using a locally deployed language model to build a persistent, incrementally expanding database of editorial board and program committee membership across disciplines. Finally, the dual-score framework, comprising BARON (Boundary-Anchored Research Outreach Network score, a strict binary metric) and HEROCON (Holistic Equilibrated Research Outreach CONstellation score, a graduated weighted metric), provides an interpretable summary in which the gap between scores serves as a structural diagnostic of inner-circle dependence. To ensure full accessibility for researchers regardless of their technical background, Citation-Constellation is available as both a command-line application and a no-code web interface (see Table 2 for links). I emphasize that the HEROCON weights are experimental defaults informed by reasoning about relationship proximity, not empirically calibrated values. I present these tools as structural diagnostics, not quality indicators. BARON and HEROCON describe where in the social graph citations originate. They should not be used for hiring, promotion, or funding decisions. 2 Positioning My Contribution The core insight, that citation patterns reflect network structure, is well-established, from Crane’s invisible colleges [ 17 ] through Wallace et al.’s network-distance analysis [ 78 ] and beyond. My contribution lies in synthesizing these disparate lines of inquiry into a unified, operational framework.

Read the complete text