Inspiration: The inspiration for this project stems from the desire to contribute to the United Nations' Sustainable Development Goals (SDGs) by leveraging technology to facilitate the research process for professors. The aim is to streamline the exploration of relevant research papers aligned with the SDGs, ultimately fostering progress towards sustainable development.
What it does: The project involves creating a tool that scans research papers to identify content related to the UN's SDGs. The tool employs the usage of an already trained ML model from "https://aurora-sdg.labs.vu.nl" to analyze and categorize the content, making it easier for professors and students in Canada to discover pertinent research in their field of interest.
How I built it: The project is built using a combination of generic data source aggregators using various techniques from web scrapping to API calls and the list is opened by design! ML model as an API call is applied to understand the content of research papers, the machine learning model is already trained to recognize and categorize content based on the SDGs. The design of classification may utilize existing libraries and frameworks and API calls.
Challenges I ran into: One of the challenges faced during the development of this project is the need build a ML model that does the classification of the text to determine whether the text falls between one or more of the specified SDGs. this involves extensive NLP, which I do not have the time to learn neither implement in the available timeframe.
Accomplishments that I'm proud of: The tool we designed in heart allows for extensibility of both data sources aggregators and classifiers, having this allows for the project to grow beyond the limitation of static design based on sole algorithm.
What I learned: Throughout the development process, I gained a deeper understanding of machine learning, I learned how to effectively apply these technologies to address real-world issues, such as facilitating research aligned with global sustainability goals.
What's next: Core code and structure is finished but further implementation needs to finish to have tool usable.