Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/bhuvantenguria/analytics_vidaya


https://github.com/bhuvantenguria/analytics_vidaya

Last synced: 1 day ago
JSON representation

Awesome Lists containing this project

README 

        
# **Welcome to the Smart Search Tool for Analytics Vidhya’s Free Courses!**  

### Your one-stop solution to effortlessly discover the most relevant courses from Analytics Vidhya’s vast collection of free resources.



---



## ✨ **Project Overview**  

Are you tired of endlessly scrolling through course lists on Analytics Vidhya, trying to find the one that matches your search? Look no further! Our **Smart Search Tool** is designed to revolutionize your course discovery experience by providing quick, relevant results through advanced search capabilities.



---



## 📚 **What is Analytics Vidhya?**  

Analytics Vidhya is one of the leading platforms offering top-notch free courses on data science, machine learning, artificial intelligence, and more. With hundreds of courses at your disposal, finding the right one can be overwhelming. Our goal is to simplify this by enabling users to search effortlessly using natural language queries.



---



## 🎯 **Why Smart Search for Free Courses?**  

- **Efficiency**: Search by title, description, or curriculum with natural language queries.  

- **Relevance**: Retrieve courses that best match your interests using semantic search techniques.  

- **Simplicity**: No more endless browsing—find exactly what you need in seconds!



---



## 🔍 **How Does It Work?**  

Here’s the journey of how we built this amazing tool:



### 1. **Data Collection & Extraction**  

We scoured Analytics Vidhya’s platform, collecting critical data such as:

- **Course Titles**  

- **Descriptions**  

- **Curriculum Breakdown**  



Using APIs and web scraping techniques, we compiled this information to form a comprehensive dataset.



### 2. **Embedding & Search Setup**  

- **Embeddings**: We used **Sentence Transformers** to create semantic embeddings of course titles and descriptions.  

- **Vector Database**: We stored these embeddings in **Pinecone** to enable fast and efficient retrieval.  

- **RAG Model**: Leveraging **LangChain (0.3.x)**, we created a Retriever-Generated Answer system that fetches the most relevant results based on cosine similarity between user queries and course embeddings.



### 3. **Deployment on Huggingface Spaces**  

- **Framework**: We built a sleek, interactive interface using **Streamlit** to ensure smooth user experience.  

- **Deployment**: The tool is live and accessible via [Huggingface Spaces](https://huggingface.co/spaces/your-username/smart-search-vidhya-courses) for public use.



---



## 🎨 **Features & Functionality**  

Here’s what makes our tool shine:  

- **Natural Language Querying**: Type your query in everyday language, and our tool understands your intent.  

- **Real-Time Suggestions**: Get the top 5 relevant course recommendations instantly.  

- **Interactive Interface**: Sleek design, smooth navigation, and a responsive layout, making your search effortless and enjoyable.



---



## 💡 **Key Libraries & Tools Used**  

- **Sentence Transformers**: For embedding course data into vector representations.  

- **LangChain**: To build and manage the RAG (Retriever-Generated Answer) system.  

- **Pinecone**: A vector database for storing and retrieving course embeddings efficiently.  

- **Streamlit**: For building the interactive web interface and deploying on Huggingface Spaces.



---



## 🌐 **Deployed Link**  

Experience the magic of our **Smart Search Tool**:  

[Visit the Tool on Huggingface Spaces](https://huggingface.co/spaces/your-username/smart-search-vidhya-courses)



---



## 📚 **How to Use**  

1. Visit the deployed tool link above.  

2. Enter your query in the search bar.  

3. Get relevant course suggestions within seconds.  

4. Click on any course to explore further details.



---



## 📝 **Methodology & Approach**  

### **Why LangChain & LlamaIndex?**  

We chose **LangChain 0.3.x** because of its powerful tools for building RAG systems that seamlessly integrate with vector databases like Pinecone. We utilized the **MiniLM model** for embedding, ensuring a balance between accuracy and efficiency.



### **Evaluation & Results**  

- **Precision & Recall**: Our search system consistently delivers relevant course suggestions, reducing search time by 70-80%.  

- **User Feedback**: Based on early feedback, we fine-tuned the RAG model, improving relevance and accuracy.



---



## 🏆 **Conclusion**  

Our Smart Search Tool for Analytics Vidhya’s Free Courses provides a seamless, interactive way to find the courses you need. With cutting-edge embedding and search techniques, it enhances the user experience, saving time and effort.  



We look forward to seeing how it can make your learning journey smoother and more efficient!