Search Science...

https://github.com/MariTheAIPsychologist/AI-course/blob/main/Search_Science_For_Our_Tomorrows.ipynb

https://vimeo.com/756063026

I attempted this AI Challenge in a team of one as a non-coder

In An Ideal World and Preserving our Scientific Legacy

If I was able to code (part of a great team, and naturally had more time alongside everyone else), to do a full scale project with The NASA Technical Report Server (NTRS) that includes these hundreds of thousands of items containing the scientific and technical information (STI) – with that focus on accessibility and discoverability of those records – for Team Humanity.

With that focus on the Cluster Data Management System (CDMS) tags and the Optical Character Recognition (OCR) and using AI to make these vast amounts of scientific knowledge, valuable data and information accessible in meaningful ways. This would include ensuring the accuracy of data of these vast amounts of scientific items and information – as an important further step in us preserving them. Especially with OCR and the potential for mistakes in data sets and errors read.

The idea would be to include detailed functionality and the option to retrieve information in rich and valuable ways – metadata, parameters, requests, queries and responses, and more. AI would give space for more intuitive information grouping, for novel ways of extrapolating data and for what the reader would want - or need. Technology features with functionalities of for instance tech notebooks that could guide a person along whilst reading/ understanding/ interpreting in even more intuitive and innovative ways - and incorporating more senses for optimal personalized learning experiences. Tech notebooks that could guide a person along whilst completing challenges and solving problems in STEM (and STEAM) - say guided along in the steps on a real-data notebook page to solve an intricate math or physics problem, yet still allowing for that people - processes - data - technology and human learning.

Going forward, we would focus on the formats of any new scientific data from the onset – say conference papers or peer reviewed journals would automatically be captured in these newly found more specific data formats – and these new data capture formats ideally used more routinely by different countries across the globe in the future.

Natural Language Processing would be used for that richness, that extrapolation of data, generating accurate and meaningful summaries, those keyword searches etc. It would also be harvested in rich and meaningful ways – gauging sentiment (with sentiment analysis) and giving information more multi-dimensional values.

Once we have preserved our science legacy with AI we would be able to access it, discover it, retrieve it, add to it, learn from it - and potentially enjoy it in innovative ways! 

These could include innovative future ways of retrieving information (and learning) - and using more of our senses during the experience too at times - see it, hear it, smell it (!), taste it (!), touch it - and even some of the more psychology concepts of that think - feel - do.

Two-way engagements with technology-newspapers, information retrieval on children's' lunchboxes, r2d2 type information companions, information in more hologram and visual-type formats, a bathtub reading us articles...

'Libraries' and information repositories for the future have the potential to be state-of-the-art and futuristic marvels of human ingenuity!

In the spirit of learning (and linked to the Hackathon Journey entry below) I used data from the NTRS - NASA Technical Reports Server - Publicly Available Content

I narrowed it down to CDMS

Then to 1992 : The Year of the Space Shuttle Endeavour for inspiration and for what I believed was a fitting tribute to this AI Challenge (...and hopefully somewhat nobel...Endeavour) :)

It was further narrowed down in scope to Conference Papers as a (sub)type

Where the first three publicly available pdf documents were used

It all started many (a few) moons ago…with a post on LinkedIn about the NASA International Space Apps Challenge 2022. Join, they said. It will be fun, they said…

First and foremost: I don’t know how to code. print(“hello, nasa!”).

This was my first ever Hackathon and I was incredibly keen to do this specific AI challenge and learn as much as I possibly could in this short SPACE of time (see what I…yes, yes you got that) :)

I am a psychologist by background with a special interest in combining my skills and love for people (most of the time) with AI and the responsible tech field. In all likelihood I will remain more on the people and AI ethics and responsible tech side in my career. I do however passionately believe in the importance of a working knowledge of the width and depth of the AI field for that solid foundation and understanding. So this Hackathon really was one small step for Man, one giant leap for Mari!

I was the only person in the whole of South Africa (to the best of my knowledge) who chose and attempted this AI challenge and that sadly left me with a team - of only me. I persevered in the spirit of "Make Space for Everybody" and that non-traditional participants were welcomed. I also absolutely love to create, explore, learn and build and am a huge Earth AND Space enthusiast.

The Hackathon also came with personal learnings and I love the Ubuntu virtues of Compassion and Humanity - we really are Team Humanity. We are so much more united. That goes for our scientific knowledge across the world as much as for our hearts!

I used Discord for the first time. I had a look at a Coursera framework to guide me through some of the Python and other concepts. I figured out how to make an Avatar and was pretty chuffed with myself. I opened a GitHub account and was practically drunk on power :) I discovered Google Colab. I dusted off my Google Drive. I looked at non-code AI proof of concepts. I searched...and searched...and searched myself :) I came, I saw, I learned loads...

The approach to this challenge was a sound: do what you can, with what you have, where you are. I made the conscious decision to really try to understand and learn the parts that I did tackle and could figure out and to keep it really simple and learn-grow-share worthy.

I would like to give a special thank you to Rene’-Peter Masoen the local Pretoria lead for all his efforts and enthusiasm. Also to the Space Agencies from around the world - to all the organisers for all your time and space spent on this Hackathon – to all the companies for their Global Offers - a special shout-out to IBM and looking forward to peruse in more detail after the Hackathon too – and of course to NASA – a heartfelt thank you. 

https://colab.research.google.com/

https://www.coursera.org/

https://discord.com/

https://www.freepik.com/

Attribution: Images by catalyststuff on Freepik 

Attribution: Image by fullvector

Attribution: Image by liuzishan on Freepik

Attribution: Image by natanealginting on Freepik

Attribution: Images by pikisuperstar on Freepik

Attribution: Image by rawpixel.com on Freepik

Attribution: Image by starline on Freepik

Attribution: Image by upklyak on Freepik

https://www.geeksforgeeks.org/python-reading-contents-of-pdf-using-ocr-optical-character-recognition/

https://www.geeksforgeeks.org/extract-text-from-pdf-file-using-python/

https://github.com/

https://ntrs.nasa.gov/

https://www.sansa.org.za/

https://www.spaceappschallenge.org/about/

https://stackoverflow.com/questions/24237111/syntax-error-invalid-syntax-for-no-apparent-reason

https://vimeo.com/

https://en.wikipedia.org/wiki/Space_Shuttle_Endeavour

https://youtu.be/RLYoEyIHL6A Google Colab Tutorial for Beginners | Get Started with Google Colab

https://zoom.us/

#humor #perseverance #pythonisfuntheysaid #fortheloveoflearning #teamhumanity