Machine Learning techniques and visualization tools for STM images at CNR-IOM labs

Master's thesis of Data Science and Scientific Computing MSc MSc. @ UniTS/SISSA

Abstract

One of the main challenges in scientific research is data management, especially in the long term. According to the European Commission data guidelines for the Horizon 2020 [1]:

Open scientific research data should be easily discoverable, accessible, assessable, intelligible, useable, and wherever possible interoperable to specific quality standards.

A clear set of principles guiding this process has been published in the FAIR Guiding Principles, which aim to render scientific data 'FAIR', that is findable, accessible, interoperable and re-usable. [2]

The Istituto Officina dei Materiali' of the Italian National Research Council (CNR-IOM) data infrastructure hosts different services for the Nanoscience Foundries & Fine Analysis NFFA-Europe project, which brings together twenty European nanoscience research laboratories with the aim to provide open access to advanced instrumentation and theory. [3] [4]

This work shows the steps taken to make FAIR a dataset of more than 100.000 scientific images obtained using Scanning Tunnelling Microscope (STM) techniques in the TASC laboratory at CNR-IOM of Trieste in the last 20 years. [5]

Repository structure

jupyter_notebooks/, contains the notebooks used for different steps of the dataset analysis
Metadata DB/, have all the code used to create the STM metadata database
TriDAS/, comprehend all the files necessary to build and run the Trieste Advance Data Services website (TriDAS)
Thesis/, files and images used in the writing of this thesis
utils/, contains .py files with functions for the STM images analysis

Grants

NEP, Nanoscience Foundries and Fine Analysis Europe|PILOT (101007417)
EOSC-Pillar Coordination and Harmonisation of National Inititiatives, Infrastructures and Data services in Central and Western Europe (857650)

References

[1] European Commision, 2016 "Guidelines on FAIR Data Management in Horizon 2020"

[2] Wilkinson et al. 2016, "The FAIR Guiding Principles for scientific data management and stewardship"

[3] Istituto Officina dei Materiali' of the Italian National Research Council, CNR-IOM

[4] Nanoscience Foundries & Fine Analysis NFFA-Europe project, NFFA-Europe

[5] Trieste Advance Data Services, TriDAS