Build Topic classifier using NLP and Python (deep learning)

Job ID: 36929929

Budget: $10 – $30 USD

** Overview:
We are seeking a skilled developer proficient in Natural Language Processing (NLP), Python, and Unsupervised Deep Learning to create an advanced topic classification system. The main focus of this project is to leverage unsupervised learning methods in NLP, including BerTopic, NMF, and LDA, to develop individual topic classification models. Additionally, we require at least two hybrid models that combine these techniques in innovative ways to achieve higher accuracy. Each hybrid model must include comprehensive evaluation methods. The dataset for this project should preferably be related to COVID-19. Furthermore, the selected developer must create a video tutorial in English or Arabic, detailing the implementation of the code and showcasing how the hybrid models operate.

** Project Description:
The selected developer will undertake the following tasks:

1. Data Collection and Preprocessing:
- Identify or acquire a relevant dataset related to COVID-19 for the topic classification task.
- Preprocess the data to ensure its suitability for unsupervised NLP tasks.

2. Unsupervised Model Implementation:
- Utilize unsupervised NLP techniques such as BerTopic, NMF, and LDA to develop individual topic
classification models.
- The learning methods for all models, including hybrid models, should be unsupervised.

3. Hybrid Model Creation:
- Devise and construct at least two hybrid models that ingeniously combine BerTopic, NMF, and LDA to improve classification accuracy.
- Each hybrid model should be unsupervised in nature.

4. Evaluation Methods:
- Implement robust evaluation metrics to assess the performance of each individual model (BerTopic, NMF, LDA).
- Perform a comprehensive evaluation for each hybrid model, with detailed analysis of their effectiveness.

5. Documentation:
- Thoroughly document the Python code, providing clear explanations of the functions and methods utilized in the project.
- Prepare a detailed report describing the methodology, experimental setup, results, and key findings.

6. Video Tutorial:
- Record a video tutorial in English or Arabic, offering a comprehensive explanation of the implementation process.
- The video tutorial should also demonstrate how the unsupervised hybrid models operate.

** Required Skills:
- Strong background in NLP, Unsupervised Deep Learning, and Python programming.
- Extensive experience with unsupervised NLP methods, including BerTopic, NMF, and LDA.
- Proficiency in designing and implementing hybrid unsupervised NLP models.
- Familiarity with popular NLP libraries, such as NLTK, spaCy, scikit-learn, and gensim.
- Ability to collect, preprocess, and work with textual data effectively.

** Deliverables:
- Well-documented Python code for the individual and hybrid topic classification models.
- Comprehensive evaluation metrics and performance analysis for all models.
- A detailed project report outlining the methodology and results.
- Video tutorial (in English or Arabic) explaining the project and demonstrating the functionality of the -unsupervised hybrid models.

** Note to Applicants:
When submitting your proposal, please provide the following:
- Your relevant experience in NLP, Unsupervised Deep Learning, and any prior work related to topic classification.
- Detailed ideas or suggestions for the two unsupervised hybrid models, and how they differ from traditional supervised approaches.
- Estimated time of completion for the entire project.
- Your preferred dataset (if any) related to COVID-19, or your plan for data collection.

We are enthusiastic about finding a developer who can showcase their expertise in Unsupervised NLP and Deep Learning and tackle this challenging project. If you have any questions or require further clarifications, please do not hesitate to reach out. We eagerly anticipate reviewing your proposals and selecting the best fit for this project.

** Important Note:
Please include the phrase "UnsupervisedNLP2023" at the beginning of your proposal to confirm that you have read and understood the job description. Proposals without this phrase will not be considered.

** Ideal Skills and Experience:
- Proficient in Python and deep learning techniques for NLP.
- Experience in building topic classifiers and achieving high accuracy.
- Ability to work with the client's dataset or assist in finding a suitable dataset.
- Willingness to accommodate specific requirements and constraints for the implementation.

If you have the necessary skills and experience, and are interested in working on this project, please submit your proposal.