MH6812: Advanced Natural Language Processing with Deep Learning
MH****6812:****AdvancedNaturalLanguageProcessing withDeep****Learning
ProjectProposalInstructions
(a) Team Size : Teams are typically composed of 3 to 5 members. However, for particularly large-scale projects, a slightly larger team size may be warranted; please confirm with the instructor for further details.
(c) Proposal: Key information to consider while determining the topic.
• Goals/Objectives : Define the objectives of your project as pertains to the scientific questions under investigation. For instance, your project could aim to explore whether a specific model or technique achieves acceptable performance on a defined task; alternatively, it might seek improvements through novel variations of existing models; for theoretical and analytical endeavors, establishing foundational principles or testing hypotheses could be central objectives. In other cases, the primary objective could be the successful development of an advanced neural architecture that demonstrates competence in addressing the problem domain. Additionally, briefly articulate the rationale behind selecting these objectives: why they are deemed important, intriguing, complex yet achievable within current technological capabilities? Furthermore, if applicable, outline any supplementary goals that may be pursued during the project's execution.
• NLP applications : The NLP applications you plan to develop for your model should be clearly defined. For each application, it is advisable to provide a detailed description, including a clear example of input and output.
Data
Neural Networks: Explain the models and techniques you will utilize. Clearly specify the components I intend to develop independently and those I will obtain from external sources. If any innovative aspect of my proposed methodology exists, ensure its acknowledgment.
• Baseline(s) : What baselines will you use to compare your model with? Make it clear if these will be implemented by you, downloaded from elsewhere, or if you will just compare with previously published scores.
• 评估：你会如何评估你的结果？请至少明确地指定一个清晰定义、数值型且自动化的定量评估度量标准，并说明你将与哪些现有的评分标准进行比较？例如，在重写或扩展现有方法时，请记录该原始方法达到的分数；如果将现有方法应用于新的任务上，请提及该新任务领域的最先进的性能水平，并预测你的方法有望超越这些水平到什么程度。此外，请具体说明你对于定性评估部分的计划。
Java Python