Please let me know where I can improve
I will be adding 2 more paragraphs about supervisor and the program
1 Para about my undergrad, 1 about my thesis process, 1 about my goals and where the inspiration comes from, latter ones about the program and the probable supervisor
Dear admissions committee,
I am writing to you with the intention of joining the PhD Program in Computer Science at George Mason University.
The primary inspiration for my career in Computer Science and Artificial Intelligence was when in high school, I learned how Google interprets and predicts words, almost like science fiction. After a period of study, programming, research in Natural Language Processing, as I prepare to advance my research in interpretable and context preserving NLP models, I believe my research interest aligns with your program's criterion.
With unwavering honesty, I will always credit my career expertise and research experience to competitive programming, as it solidified my problem solving abilities, something that transcends the boundary of mindless coding and helps approach problems creatively. This has polished my journey through courses like Object Oriented Programming, Data Structure and Algorithms, even Artificial Intelligence and Neural Networks courses. On the other hand, mathematics and statistics courses were a crucial part of my journey in AI. My knowledge in probability, linear equations, derivatives and vectors helped tremendously in gaining logical understanding of something seemingly magical. Online courses, notably lectures from Dr. Andrew Ng strengthened my mathematical foundations of AI, while I used my programmatic expertise to implement AI models. The entirety of my undergraduate experience has surely added up to my research prowess.
I have always wanted to create something with significant social impact and creativity. Having chosen Natural Language Processing as my primary research area, I focused on polarizing sentences, i.e., Sarcasm or Irony Classification. My choice was inspired by the surge in racial profiling shrouded in Sarcasm, which made counter-measures ineffective. It also wrongly flagged harmless Sarcasm, hindering freedom of expression. Having read literature that heavily focused on model accuracy, I shifted my focus instead toward resource-constrained, simplified classification models. A significant hurdle here was using singular models, like 1-Dimensional Convolution Neural Network, which was overfitting with fast convergence, and a Long Short Term Memory, which was more accurate but converged slowly. So, using a hybrid model, layering a 1D-CNN for data extraction with LSTM, I achieved adequate convergence metrics while maintaining great accuracy, earning the highest GPA in my thesis and a publication.
As a Software Developer I work on a Microfinance Solution that enables the goodwill of one-third of the country's population. The fact that my work can impact people's lives, makes me work with grit and focus. This has fixed my goal towards improving our society through research and implementations.
My current goal is to work with interpretable and efficient NLP models capable of wide contextual relation in interaction tasks. I believe modern LLMs can get biased or ignore sentiments due to limited contextual relation, making interactions seem unnatural. Inconsistencies in cultural sensitivity can result in offensive text generation. Parallely, complex LLM models consume large resources, affecting the environment with carbon footprint. My research in these areas can help build culturally conscious and efficient AI which can impact Education, Health Industry and Bias Mitigation, while having minimal resource consumption.
In this regard, I believe Professor X's work in blah blah blah is very inspiring.
I genuinely admire the research and collaboration culture of your program. I believe knowledge can only increase through spreading, and this program strives on that. Being from a South-Asian Descent, the diverse faculty and student core feels like fresh air. This will enable me to concretely build AI Solutions that are culturally sensible and generate diverse ideas.
While my mind was set on my career prospects when choosing this program, what captivated me was the program's interdisciplinary approach, making it possible to collaborate with other fields of study. Such collaboration can have a widespread impact in Psychology, Social Science, Medical Enhancements etc.. The program's facilitation of interdisciplinary collaboration will surely generate broad scale ideas for my work to have a positive social impact, which is my life's goal.
I will be adding 2 more paragraphs about supervisor and the program
1 Para about my undergrad, 1 about my thesis process, 1 about my goals and where the inspiration comes from, latter ones about the program and the probable supervisor
Dear admissions committee,
I am writing to you with the intention of joining the PhD Program in Computer Science at George Mason University.
The primary inspiration for my career in Computer Science and Artificial Intelligence was when in high school, I learned how Google interprets and predicts words, almost like science fiction. After a period of study, programming, research in Natural Language Processing, as I prepare to advance my research in interpretable and context preserving NLP models, I believe my research interest aligns with your program's criterion.
With unwavering honesty, I will always credit my career expertise and research experience to competitive programming, as it solidified my problem solving abilities, something that transcends the boundary of mindless coding and helps approach problems creatively. This has polished my journey through courses like Object Oriented Programming, Data Structure and Algorithms, even Artificial Intelligence and Neural Networks courses. On the other hand, mathematics and statistics courses were a crucial part of my journey in AI. My knowledge in probability, linear equations, derivatives and vectors helped tremendously in gaining logical understanding of something seemingly magical. Online courses, notably lectures from Dr. Andrew Ng strengthened my mathematical foundations of AI, while I used my programmatic expertise to implement AI models. The entirety of my undergraduate experience has surely added up to my research prowess.
I have always wanted to create something with significant social impact and creativity. Having chosen Natural Language Processing as my primary research area, I focused on polarizing sentences, i.e., Sarcasm or Irony Classification. My choice was inspired by the surge in racial profiling shrouded in Sarcasm, which made counter-measures ineffective. It also wrongly flagged harmless Sarcasm, hindering freedom of expression. Having read literature that heavily focused on model accuracy, I shifted my focus instead toward resource-constrained, simplified classification models. A significant hurdle here was using singular models, like 1-Dimensional Convolution Neural Network, which was overfitting with fast convergence, and a Long Short Term Memory, which was more accurate but converged slowly. So, using a hybrid model, layering a 1D-CNN for data extraction with LSTM, I achieved adequate convergence metrics while maintaining great accuracy, earning the highest GPA in my thesis and a publication.
As a Software Developer I work on a Microfinance Solution that enables the goodwill of one-third of the country's population. The fact that my work can impact people's lives, makes me work with grit and focus. This has fixed my goal towards improving our society through research and implementations.
My current goal is to work with interpretable and efficient NLP models capable of wide contextual relation in interaction tasks. I believe modern LLMs can get biased or ignore sentiments due to limited contextual relation, making interactions seem unnatural. Inconsistencies in cultural sensitivity can result in offensive text generation. Parallely, complex LLM models consume large resources, affecting the environment with carbon footprint. My research in these areas can help build culturally conscious and efficient AI which can impact Education, Health Industry and Bias Mitigation, while having minimal resource consumption.
In this regard, I believe Professor X's work in blah blah blah is very inspiring.
I genuinely admire the research and collaboration culture of your program. I believe knowledge can only increase through spreading, and this program strives on that. Being from a South-Asian Descent, the diverse faculty and student core feels like fresh air. This will enable me to concretely build AI Solutions that are culturally sensible and generate diverse ideas.
While my mind was set on my career prospects when choosing this program, what captivated me was the program's interdisciplinary approach, making it possible to collaborate with other fields of study. Such collaboration can have a widespread impact in Psychology, Social Science, Medical Enhancements etc.. The program's facilitation of interdisciplinary collaboration will surely generate broad scale ideas for my work to have a positive social impact, which is my life's goal.
