A Semester of Research
All of this comes together for a Fake Detection Network, or FaDe-Net. In order to validate and benchmark our results, we performed a series of metric tests on the MIT Fake News Dataset1 and achieved a precision rate of 90.31% as well as an accuracy rate of 88.56%. Our experiments indicate the model is more effective at fake news detection. However, this model still has a significant amount of room to improve in future research.
No. The naivety of that statement still makes me cringe to this day. That was still during my senior year of high school where I was conducting a research project as part of an Advanced Placement (AP) Research class. Although the research wasn’t the most clear-cut, that class still taught me an immense amount of valuable information that I still use on a frequent basis.
Being an actual research class where human subjects can come into play, we went through and performed every section of the research process. That means writing IRBs, gaining approval, finding a sponsor, just to list a few. Albeit this was stressful at times, the beauty of it was seeing the results that eventually came to fruition. In addition to that, I strongly believe that it has set me up for future success in college level research.
I had two significant takeaways from that high school class. The first takeaway was being able to read and summarize immense amounts of literature. To be fair, literature reviews are despised by many people. No one wants to be sat down, organize and summarize that text, and then present it only to be told it isn’t enough or more needs to be read. Through this mundane process, I learned how to effectively read academic articles efficiently which allowed me to glean the vital information in an article. This resource here is something I recommend to everyone who reads academic papers frequently. I usually apply the techniques found in the paper if I need to gain a high level overview of a specific industry or topic that I am interested in, or my research is headed towards—not needing to spend more than a day or two reading papers for extended periods of time.
The second takeaway I had from this was learning how to develop a sense of academic independence. By academic independence, I don’t mean someone who is able to study by themselves. I mean being able to explore a variety of topics without professional guidance; taking an inkling of a topic and running with it; or developing the curiosity and fearlessness to try new things. There is no “cookie cutter” method to conducting research. Everyone has their own style and different fields vary significantly. One may enjoy reading literature while another may loathe it. However, no matter what field of research you are conducting, many of them have these aforementioned traits in common. Even with my naivety of my works in high school, I continued to sow that seed which has lead to my work currently in college.
Fast-forward to today, I’m currently doing work in a field I am extremely passionate about: natural language processing (NLP) and their applications in different situations. Even with what I learned in high school, nothing prepared me for the fast pace of college research as a whole.
Ecstasy. The first feeling when I realized that I was accepted to conduct research in a lab working with machine learning (ML) and NLP systems. I felt that it was as if I am following the traditional path many academics took: conducting researchers, going to conferences, and presenting results. The decision was a no-brainer, I immediately accepted the offer and started officially in late January.
However, even in such a setting where they claim to “teach you along the way,” a level of technical competency was still expected and anything you didn’t know you were expected to learn in your free time. There were minimal amounts of hand-holding and required a-lot of external commitment outside traditional hours. Although I struggled at times, the more I enjoyed it at the time, I found that the more I learned. What I was once thought was a complex AP Research project finished in high school was dwarfed by the immense complexity, span, and technical acumen required for some projects.
One thing I observed that having a high level of organization is directly correlated with success. A habit I started was writing and compiling weekly reports which consisted of various charts, statistics, and work accomplished for that current week. At first, they proved to be tedious, but soon I learned to effectively document my work. This became extremely critical later on when a final report was needed, or we needed to present a poster at a conference.
In addition to that, persistency is also critical to success in research as well. Many times, the number of failures will exceed the number of successes. In fact, it isn’t just many, but it is a guarantee. While at some times it may be disheartening, you just have to not give up and keep pushing forward. I can’t count the amount of times I’ve had a model fail to compile at the last moment after 14 hours of training or even a miscalculation of resources. The failure doesn’t deter me but instead entices me: “what if it worked?”
After talking about all of this, I’ll leave with some of my tips for those who want to get involved in undergraduate research. One of the best ways is to reach out! Don’t be afraid to ask a professor who has research that interests you. It never hurts to send out an email, You either try for it or you fail, and it’s better to try than fail. Another good way to get involved is to join various labs that offer research opportunities. On campus, there are many labs structured with multiple Principal Investigators (PIs) and research projects available to your interest. These are also available to a huge variety of majors as well! At the end of the day, it’s just your interest and passion that will drive the most success. Finally, from my observations, what I strongly believe in is that it is your own passion and enthusiasm that will drive success within research. Don’t be afraid to go the extra mile on a project or checkpoint, your work would always be recognized at the end of the day. Sometimes, it’s not about
As I would say, if you are unsure about such things, just full send.