High-Level Project Summary
High-Level Project Summary*Provide a high-level summary of your project. What did you develop? How does it "solve" the challenge? Why is it important?We created an AI application that accesses NTRS database, calls the API, generates text analytics, provides a brief summary of each article, and displays a histogram for most repeated words using NLP.Using Extractive Summarization Approach, a brief summary of each article is generated; the three main steps of making each summary are: 1- Generation of an Intermediate Representation2- Assign a score to each sentence (Clarifying sentence importance)3- Select Sentences for the Summary ( The most relevant k(30%) sentences are selected )
Link to Final Project
Link to Project "Demo"
Detailed Project Description
.
Our project is starting with a simulation of the NTRS website as a desktop application to emphasize the feature of Summarizing and Analyzing the Papers on the NTRS database to make them more accessible and easy to discover what the content it takes without much reading or need to read the whole paper and that to ease the life of researchers as well as make the Space Science available to everyone.
The application is working as follows:
First, you sign in to the application and check if you have permission to open private data on NTRS or not. Secondly, we have simulated the search bar and filter at a very low scale to be able to search for needed articles. All of that UI was made by PyQt5. All the articles/papers are stored as a pdf file in the corpus folder in the project. When searching the application is fetching in the corpus directory using the “os” package in python then reads the pdf files in python as a string and calls a summarizing API that we have built using the NLP algorithm using Spacy with the extractive summarization and using flask package to host the API in local server. After that, it returns the summarization of each article which appear under its name in the search result. The benefit of this project is to ease the accessibility to the papers/content to decide whether to read it or not. We intend to add more features to it such as the Text analysis which displays the keywords and most used words as well as graphs of that data. Moreover, we intend to join it with more advanced websites/applications such as the NTRS with new features added will be designed. All the project was made using python language with the previously mentioned packages.
Space Agency Data
We use The NASA Technical Report Server (NTRS) that includes hundreds of thousands of items containing scientific and technical information (STI) created or funded by NASA. by utilizing it's documents in form of pdf files to summarize them and it inspired us by enabling searches of this large NTRS database, potential users to read the summary of these scientific papers to attract them and encourage them to get more details -
Hackathon Journey
.This hackathon has been totally a wonderful experience. We as a team have together achieved a great project that otherwise we would not imagine doing by the start of the hackathon. Everyone has impressively excelled at their tasks and gone beyond expectations. In addition, we have been solidifying each other with great support. All of us have shown to be nothing but true and compassionate friends.
We made our decision of choosing this challenge, CAN AI PRESERVE OUR LEGACY? by reading, all challenges, sharing opinions, outlining skills, and deciding our tendency
We have managed to go through all holdbacks by kindly and collaboratively sharing knowledge, transferring skills, checking progress and performance, clearly expressing opinions, and actively spreading a positive attitude.
I would love to give my sincere and genuine thanks to all my team members. We have been a solid body throughout the whole journey from choosing the challenge to implementing and submitting the final project. We together managed to overcome all the obstacles we faced. Besides all the pressure and the rapid pace of the hackathon, we all stood together as a single unit, collaborating and reassuring each other of our great capabilities.
References
- .https://ntrs.nasa.gov/
- https://ntrs.nasa.gov/api/openapi/
- https://ntrs.nasa.gov/api/citations/20000025197/downloads/20000025197.pdf
- https://sti.nasa.gov/docs/STI_Open_API_Documentation_20210426.pdf
- https://www.youtube.com/watch?v=dIUTsFT2MeQ
- https://code.visualstudio.com/docs/python/tutorial-flask
- https://iq.opengenus.org/text-summarization-techniques/#:~:text=Following%20are%20the%20text%20summarization%20techniques%3A%201%20Luhn%27s,6%20TextRank%207%20Reduction%208%20Latent%20Semantic%20Analysis
- https://www.youtube.com/watch?v=fNxaJsNG3-s&list=PLQY2H8rRoyvzDbLUZkbudP-MFQZwNmU4S
- https://pypi.org/project/pyqt5ac/
- https://sti.nasa.gov/docs/OpenAPI-Data-Dictionary-062021.pdf
Tags
#Software , #AI ,#algorithm ,#NLP