Recommendation Model for Women in STEM Opportunities

Inspiration

Science, Technology, Engineering, and Math, collectively known as STEM, are the four primary academic disciplines currently responsible for driving the entire world’s economy and maintaining our general well-being. Students who receive a quality STEM (Science Technology Engineering Mathematics) education are primed to become the next generation of innovators as STEM coursework challenges students to think critically and produce their own solutions. Employment in STEM occupations is projected to grow 8.8% by 2028, meanwhile, non-STEM occupations will only grow by 5%. Women have made tremendous progress in education and the workplace during the past 50 years. Even in historically male fields such as business, law, and medicine, women have made impressive gains. In scientific areas, however, women’s educational gains have been less dramatic, and their progress in the workplace is still slower.

What it does

With this project, we aspire to develop a strong recommendation model which will create an environment of encouragement that can disrupt negative stereotypes about women’s capacity in these demanding fields. By supporting the development of girls’ confidence in their ability to learn math and science, we help motivate interest in these fields. Women’s educational progress should be celebrated, yet more work is needed to ensure that women and girls have full access to educational and employment opportunities in STEM. This model will also mention connection with women senior leaders, who have acquired a respectable position in the field of STEM, as mentors to the women in the same expertise. With all this features, we anticipate that this model will serve its purpose and will help women to thrive and contribute in the STEM field.

How we built it

Our first step towards the project was to collect the data available. We have scraped data from Linkedin by using specific search term as "Her/She" to gather women's data. Subsequently, we have gathered data jobs and certification data from multiple sites. As the data received from the sites was raw data which consisted of profiles not related to STEM courses, therefore we filtered the CSV using pandas lib and searched keywords. For our recommendation model, we have used TigerGraph's DataScience library to create a supervised recommendation engine. Within TigerGraph's DS Library, we have used similarity algorithms which assigns a score to a pair of vertices based on how similar they are. In this particular scenario, it usually involves seeing if two individuals have the same or similar skillsets. Similarity algorithms used here is Cosine similarity.

Challenges we ran into

The main challenge in our recommendation model was to prepare an appropriate database which will be suitable in our model. The main road blocker was to gather the dataset by scrapping the LinkedIn, as there was no gender-based filtering criteria. We have to filter it out, on the basis of search term like “She/her” which gave us a subset of women’s data present on linked in.

Accomplishments that we're proud of

We were able to create the platform for such women, who are aspiring career in STEM. This was achieved by connecting them to the respective women who have achieved excellence in that field. Also, these aspiring women will be recommended certifications and jobs which will help them to excel in their domain. This was all achieved with the help of TigerGraph Recommendation Engine which uses Cosine Similarity.