Publications SISSA
URI permanente per questa collezione
Sfogliare
Immissioni recenti
1 - 5 di 14352
-
PubblicazioneData Curation for Optimizing Molecular Beam Epitaxial Growth of III-V Semiconductor Samples( 2025-05-27)This thesis presents a case study on the FAIRification of scientific data within the context of the NFFA-DI initiative, which promotes open science and data standardization in materials science and nanotechnology. The goal was to enhance the quality, usability, and long-term accessibility of legacy laboratory data collected at the HMMBE Laboratory, a molecular beam epitaxy synthesis facility. To achieve this, a complete workflow was developed to convert fragmented, unstructured text files into structured, interoperable datasets using the NeXus format, a hierarchical standard based on HDF5. A new application definition and two supporting base classes were created to model the MBE deposition process. Parsing functions were implemented to extract, transform, and enrich data from heterogeneous sources, resulting in fully structured NeXus files. Consequently, the converted files were validated both in the structure and content. The transformed data were then uploaded in a local NOMAD repository. A custom plugin and dedicated NOMAD app were developed to enable proper ingestion, indexing, and visualization of the new data format. The outcome demonstrates how isolated experimental records can be converted into reusable and interoperable scientific assets, contributing to the advancement of FAIR data practices and supporting broader goals in open, collaborative research.
-
PubblicazioneImplementation of FAIR by design principles for data acquisition and storage at CNR-IFN Milano( 2025-05-28)This thesis presents the work conducted during an internship at CNR-IFN@MI within the UDynI research group. Following an initial contextualization of Open Science principles and the research activities at the CNR-IFN laboratories, the core of the project, the design and implementation of a digital data management infrastructure aligned with the FAIR principles, is described in detail. The internship resulted in the development of two software components. The first, udyninexus, is a Python package designed to facilitate the creation of NeXus-compliant data files, developed in conjunction with the application definition chosen for the laboratories. This enables a smoother integration of standardized, structured data into the lab’s acquisition workflows. The second component, LabLogbook, is a web-based application that allows users to digitally manage experiments, samples, and measurement sessions, offering a centralized and version-controlled environment for recording experimental activities. Together, these components establish a scalable and maintainable foundation for managing experimental data within the lab, fostering improved documentation, long-term accessibility, and reproducibility of scientific out
-
PubblicazioneImplementation of FAIR data in Transient Absorption Spectroscopy and Photoluminescence Spectroscopy( 2025-05-27)This thesis explores the implementation of FAIR data principles — Findability, Accessibility, Interoperability, and Reusability — within the context of the spectroscopy experiments conducted at the EuroFEL Support Laboratory. The project focuses on the adoption of the NeXus data format to adapt the output data from the Femtosecond Transient Absorption Spectroscopy and Photoluminescence Spectroscopy beamlines. For the transient absorption setup, the existing LabVIEW acquisition system was adapted to include the required metadata structures defined by the NXoptical_spectroscopy application definition. In parallel, a graphical user interface was developed to convert proprietary .spc files into NeXus format for photoluminescence measurements. All files generated were tested for cross-platform interoperability and FAIR compliance using open-source visualization tools. This work contributes to the standardization and transparency of scientific data workflows, supporting long-term goals of Open Science as outlined by international ini
-
PubblicazioneImplementation of a pipeline for collecting, ingesting and transforming data into standard formats for the LAME FIB-SEM( 2025-05-27)This thesis presents the development of a FAIR-by-design data management platform for the Laboratory of Electron Microscopy (LAME) at Area Science Park in Trieste, Italy. LAME supports high-resolution materials research using techniques such as Scanning Electron Microscopy (SEM), Focused Ion Beam-SEM (FIB-SEM), and Scanning Transmission Electron Microscopy (STEM), generating large volumes of complex data that require structured, interoperable management. The platform streamlines data acquisition, metadata capture, validation, and publication in line with the FAIR (Findable, Accessible, Interoperable, Reusable) principles. Built using Django, it enables researchers to upload structured NeXusformat data through an intuitive web interface. All components are deployed within ORFEO, the centralized data center at Area Science Park, which hosts MinIO for object storage, Authentik for secure authentication, and NOMAD Oasis for FAIR-compliant internal publishing under the NFFA-DI (Nano Foundries and Fine Analysis – Digital Infrastructure) framework. Datasets are also integrated into OFED (Overarching FAIR Ecosystem for Data), ensuring standardized, traceable, cross-institutional access. The platform is aligned with NFFA-Europe standards via tools like MetaRepo and will soon be publicly hosted on ORFEO. The underlying codebase and documentation will be made openly available on GitHub to support reuse and adoption by other scientific facilities.
-
PubblicazioneThe role of population structure in computations through neural dynamics( 2022)Neural computations are currently investigated using two separate approaches: sorting neurons into functional subpopulations or examining the low-dimensional dynamics of collective activity. Whether and how these two aspects interact to shape computations is currently unclear. Using a novel approach to extract computational mechanisms from networks trained on neuroscience tasks, here we show that the dimensionality of the dynamics and subpopulation structure play fundamentally complementary roles. Although various tasks can be implemented by increasing the dimensionality in networks with fully random population structure, flexible input–output mappings instead require a non-random population structure that can be described in terms of multiple subpopulations. Our analyses revealed that such a subpopulation structure enables flexible computations through a mechanism based on gain-controlled modulations that flexibly shape the collective dynamics. Our results lead to task-specific predictions for the structure of neural selectivity, for inactivation experiments and for the implication of different neurons in multi-tasking.