Electrospinning-Data.org: A FAIR, Structured Knowledge Resource for Nanofiber Fabrication
Mehrab Mahdian, Ferenc Ender, Tamas Pardy · 2026
Read Electrospinning-Data.org: A FAIR, Structured Knowledge Resource for Nanofiber Fabrication by Mehrab Mahdian, Ferenc Ender, Tamas Pardy (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-2603-27841.txt.
Introduction
Mahdian and co-authors launch Electrospinning-Data.org, a shared nanofiber data hub. Fiber traits hinge on many linked lab settings. Records today stay scattered and skew toward successes. The platform aligns with FAIR data principles. A process-structure-property model organizes each experiment. Controlled vocabulary tags keep entries machine-readable. Two-stage moderation mixes auto checks with expert review. Failure-aware records support reuse across labs.
Catalog record
| CREATOR | Mehrab Mahdian, Ferenc Ender, Tamas Pardy |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40685 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.27841
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Electrospinning-Data.org: A FAIR, Structured Knowledge Resource for Nanofiber Fabrication Mehrab Mahdian, Ferenc Ender, Tamas Pardy Abstract 1 Introduction Electrospinning is a widely adopted nanofabrication technique for producing continuous fibers from polymer solutions. These fibers typically exhibit nanoscale diameters and high surface-to-volume ratios, making their structural characteristics central to their functional performance. Its versatility has enabled applications across tissue engineering Pisani2021DesignOfExperiment , drug delivery Sill2008DrugDelivery , filtration Shao2024BimodalNanofibers , energy storage Doroudkhani2025NiOMn3O4 , and protective textiles Wang2025MultiStructuredNanofibers , among many other domains. Across these applications, performance is governed by both functional and mechanical properties, which are strongly linked to structural features of the nanofiber mat, including fiber diameter, diameter distribution, porosity, and the presence of defects. These structural features—and consequently material performance—are highly sensitive to the interplay between solution properties, process parameters, and environmental conditions, creating a complex and multidimensional experimental landscape in which small variations in processing conditions can lead to significant changes in fiber morphology Deitzel2001ProcessingVariables ; Son2004EffectsSolution ; Haghi2007Trends ; Zong2002StructureProcess ; Casper2004ControllingMorphology . Optimizing electrospinning processes for targeted applications thus requires systematic exploration of this high-dimensional parameter space, which is often conducted through trial-and-error experimentation Li2004Electrospinning . This intrinsic complexity has led to the generation of vast amounts of experimental data across laboratories worldwide, as researchers explore diverse material systems and processing windows. Nonetheless, the data remain largely fragmented and underutilized due to inconsistent reporting and a pervasive publication bias toward successful outcomes Ramakrishna2005ElectrospinningNanofibers ; Xue2019ElectrospinningReview ; Mahdian2025TechRxiv . This success-biased reporting obscures the true boundaries of feasible processing windows and limits reproducibility, as researchers lack access to negative results that are important for understanding instability mechanisms Mahdian2025TechRxiv ; Fanelli2012NegativeResults . From a data-centric perspective, the absence of failure information introduces systematic bias into datasets, reducing their reliability for statistical analysis and machine learning applications. A systematic analysis of the electrospinning literature confirms substantial reporting inconsistency and data fragmentation. In our prior review of 312 screened papers, only 137 ( \approx 44\% ) contained sufficiently reported quantitative parameters for structured extraction, yielding 1508 experimental records Mahdian2025TechRxiv . After filtering for completeness and usability, only \approx 56\% of collected data were suitable for quantitative analysis. Parameter variability across studies ranged between 30–50%, even for commonly used materials and flat-collector setups, indicating substantial heterogeneity in experimental reporting and control. Environmental parameters are particularly underreported. Although temperature and humidity critically affect morphology, many records lack these values or report them without controlled ranges. Furthermore, instability and failure descriptors are rarely standardized in the electrospinning literature, while morphology beyond mean fiber diameter is typically described using subjective terminology Mahdian2025TechRxiv . Consequently, much of the field’s experimental knowledge remains embedded in narrative reports rather than structured, machine-readable records. This limits reproducibility, systematic cross-study comparison, and the development of data-driven models. These limitations are particularly consequential in the context of machine learning, where model performance is fundamentally constrained by the structure, completeness, and representativeness of the underlying training data Halevy2009UnreasonableEffectiveness — none of which the current reporting landscape consistently provides. Meeting this need requires more than a static database. The convergence of experimental materials science, FAIR (Findable, Accessible, Interoperable, and Reusable) data engineering, and machine learning is contributing to a more data-centric mode of materials research Gray2009FourthParadigm ; Wilkinson2016FAIR ; Agrawal2016MaterialsInformatics , in which systematically curated datasets become an important scientific asset and facilitate analyses that are difficult to carry out across isolated studies. In electrospinning, this shift calls for a live, governed infrastructure that can organize dispersed experimental reports into a coherent, community-oriented knowledge resource. In practice, the need for such infrastructure becomes evident across the breadth of electrospinning research. Any researcher working to optimize fiber morphology, porosity, or mechanical properties must navigate a high-dimensional parameter space—spanning material composition, solution properties, process parameters, and environmental conditions—where each variable interacts in ways that are rarely captured in a single study. Consider, for instance, a researcher developing polyvinyl alcohol (PVA) nanofiber mats for filtration: identifying viable processing windows requires laborious extraction from fragmented literature, where negative results and instability data are seldom reported, forcing optimization to rely heavily on trial-and-error experimentation. This challenge is not unique to filtration applications—it extends equally to biomedical scaffolds, energy storage electrodes, and functional textiles. This limitation is particularly acute in the era of machine learning, where predictive models may help identify stable processing windows and target morphologies but remain constrained by the lack of diverse, failure-inclusive training data. A centralized, structured platform can enable researchers across domains to explore and analyze prior experiments—including unsuccessful attempts—and systematically relate processing conditions to reproducible morphological outcomes. Existing efforts to organize experimental materials data span a wide range of scopes, from domain-specific datasets to large general-purpose materials data infrastructures. Within electrospinning, material-focused datasets such as FEAD Sarma2023ElectrospunFEAD provide curated records for a specific polymer system (i.e., PVDF), while our earlier work, Cogni-e-SpinDB 1.0 Mahdian2026CogniESpinDBArticle ; Mahdian2025CogniESpinDB , introduced a broader, machine-readable dataset of 809 electrospinning parameter records across multiple polymers. Beyond the electrospinning domain, general materials science platforms—including the Materials Project Jain2013MaterialsProject , NOMAD Draxl2019NOMAD , and the Materials Data Facility (MDF) Blaiszik2019MDF have established robust infrastructures for sharing computational and experimental materials data at scale (Table 1 ). However, these resources differ fundamentally in their data models, update mechanisms, and domain assumptions. General materials platforms prioritize crystallographic, electronic, or thermodynamic properties and are not designed to capture process-driven experimental workflows, qualitative morphology descriptors, or failure and instability information intrinsic to electrospinning experiments. Conversely, electrospinning-specific datasets, while valuable, are typically released as static archives and lack continuous data ingestion, consistent morphology representation, structured failure representation, image integration, and expert-governed curation. Electrospinning-Data.org is designed to address this gap by providing a domain-specific infrastructure for structuring electrospinning experiments within a unified process–structure–property framework. To our knowledge, there is currently no continuously updated, failure-aware infrastructure that combines structured experimental representation, governed curation, and support for both successful and unsuccessful outcomes within a single platform. Table 1: Comparison of Electrospinning-Data.org with existing electrospinning and materials science data resources. Electrospinning-Data.org Cogni-e-SpinDB FEAD Materials Project NOMAD MDF Domain scope Electrospinning (multi-polymer) Electrospinning (multi-polymer) Electrospinning (PVDF only) Inorganic materials Computational materials General materials Data representation Structured records + linked images Structured tabular dataset Structured tabular dataset Computed & measured properties Structured repository Heterogeneous datasets Process–structure–property framework Yes (unified schema) Partial Partial No No No Failure & instability recording Yes (structured, schema-level) Partial Partial N/A No No Morphology controlled vocabulary Yes (Cogni-EMCV) No No N/A N/A N/A Image evidence integration Yes (linked SEM images) No No No Partial Partial FAIR alignment Explicit Explicit Limited Explicit Explicit Explicit Provenance & versioning Record-level provenance, versioned releases Dataset-level versioning No Versioned releases Partial Partial Governance & update model Expert-moderated, continuous ingestion Centralized, static Centralized, static Centralized, periodic Community-driven Community-driven Contribution workflow Structured submission + automated validation + expert review N/A N/A Curated internally Open upload Open upload Platform role Live curated data platform Static dataset Static dataset Property database Data repository Data hosting While Cogni-e-SpinDB 1.0 established the feasibility of structuring electrospinning experiments into a machine-readable dataset, it remains a static resource with fixed scope and limited capacity for continuous expansion. As such, it does not address key challenges in the field, including the ongoing integration of newly published data, the systematic inclusion