Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due
Project Description: Analyzing Nvidia’s Stock Movements Using NLP (5 Bonus Points)
Course: ECOM217 — Large Language Models and Textual Analysis in Finance
1 Overview
This project involves conducting an industry-level report that investigates the factors boosting Nvidia’s stock price rally in the AI industry before January 27, 2025, and explores the significant decline in the stock price on January 27, 2025, following the release of DeepSeek. The analysis leverages textual analysis techniques and large language models (LLMs) alongside statistical tools.
2 Deliverables
The final submission should consist of:
1. Presentation Slides: A slide deck summarizing the research approach, methodology, key findings, and visualizations. Presentation Day: Tuesday, April 8th, 2025
2. Code Script: A well-documented Python script or Jupyter Notebook detailing data col- lection, preprocessing, analysis, and visualization. Submission Deadline: Tuesday, April 22nd, 2025
3. Analysis Report: A comprehensive report (minimum 3500 words, excluding tables, graphs, and appendices) that details the methodology, analysis, results, and conclusions. Submission Deadline: Tuesday, April 22nd, 2025
3 Project Tasks and Requirements
3.1 Part 1: Analyzing the Pre-Rally Events
• Data Collection:
— Collect textual data from financial news outlets, press releases, social media, and analyst reports.
— Focus on articles and posts discussing Nvidia, AI breakthroughs, and related industry events.
— Cover the period leading up to January 27, 2025.
• Data Preprocessing:
— Clean the data by removing noise (e.g., stopwords, punctuation, URLs) and tokenize the texts.
— Normalize the text using stemming or lemmatization.
— Extract features using techniques such as Bag-of-Words, TF-IDF, or word embeddings.
• Exploratory Analysis:
— Identify key events via keyword frequency analysis and clustering/topic modeling (e.g., Latent Dirichlet Allocation).
— Conduct sentiment analysis using models such as Na¨ıve Bayes or logistic regression.
— Visualize sentiment trends over time.
• Interpretation:
— Analyze how the identified events influenced market sentiment and Nvidia’s stock rally.
— Employ regression or correlation analyses to link news events and stock price movements.
3.2 Part 2: Investigating the DeepSeek Event on January 27, 2025
• Data Collection:
— Gather pre- and post-event data, including analyst forecasts, financial news, and social media commentary focused on DeepSeek.
• Comparative Analysis:
— Compare sentiment distributions before and after January 27, 2025.
— Extract and aggregate analyst forecasts to compare market expectations.
— Determine if the impact is short-term or indicative of a longer-term shift using time-series analysis and hypothesis testing.
• Visualization and Modeling:
— Use time-series plots, sentiment histograms, and comparison tables.
— Apply change-point detection methods to validate the impact’s timing and magnitude.
4 Timeline and Work Allocation (1-Month Duration)
• Week 1: Define research questions, set up data collection pipelines, and plan the preprocess- ing steps.
• Week 2: Complete data cleaning and preprocessing; perform exploratory analysis and initial sentiment/topic analysis.
• Presentation Day: Tuesday, April 8th, 2025
• Week 3: Focus on the DeepSeek event analysis; collect additional data, run comparative analyses, and perform statistical tests.
• Week 4: Finalize analyses and visualizations; compile the code script and slide deck; write and revise the final report.
• Submission DDL: Tuesday, April 22nd, 2025
5 Evaluation Criteria
• Analytical Rigor: Systematic data collection, preprocessing, and the proper application of NLP and statistical techniques.
• Methodological Clarity: Well-documented code, reproducible results, and a logical flow in the analysis report linking textual insights to financial outcomes.
• Communication: Professional-quality slides and a comprehensive report discussing method- ology, analysis, results, limitations, and recommendations.
• Collaboration: Effective teamwork among 5-8 group members with clear task division.
6 Final Notes
Students are encouraged to leverage course materials, including lectures on NLP fundamentals, sentiment analysis, vector space models, and statistical techniques, to guide their work. The project should not only demonstrate technical proficiency but also critically evaluate the broader implications of textual analysis in financial decision-making.