🔍 Mini Search Engine in Java

A console-based Mini Search Engine built using Core Java, implementing key concepts from Data Structures, Algorithms, and Information Retrieval systems.

The project supports fast keyword-based search, partial-word queries, and relevance-based ranking using TF-IDF, similar to how real-world search engines work internally.

---

🚀 Features

• ✅ Keyword-based document search
• ✅ Supports partially typed words (prefix search)
• ✅ Efficient relevance ranking using TF-IDF
• ✅ Returns Top 5 most relevant documents
• ✅ Highlights matched keywords in results
• ✅ Optimized for performance (no combinatorial explosion)
• ✅ Clean, modular, interview-ready code

---

🧠 How the Search Engine Works

1. Documents are indexed using an Inverted Index
2. User query is tokenized
3. Partial words are expanded using a Trie
   • Example: ba → backend
4. Expanded keywords are merged into a single query
5. TF-IDF scores are computed for each document
6. Documents are ranked and Top 5 results are returned

---

⚡ Time Complexity Optimization (Important)

❌ Naive Approach (Not Used)

• Generate all combinations of prefix-expanded words
• Leads to exponential time complexity

O(k^p × D × log D)

✅ Optimized Approach (Used in This Project)

• No combinatorial query generation
• Single-pass TF-IDF computation using all expanded keywords
• Uses a Priority Queue to keep only Top 5 results

Final Complexity:

O(T × avgDocs)

Where:

• T = number of expanded query terms
• avgDocs = average documents per term

➡️ This makes the engine scalable and production-like.

---

🧱 Data Structures Used

Data Structure	Purpose
HashMap	Inverted Index, document storage
Trie	Prefix-based word expansion
PriorityQueue	Ranking top search results
Set	Remove duplicate keywords
List	Store documents and results

---

🧮 Algorithms Used

• Inverted Indexing
• TF-IDF (Term Frequency – Inverse Document Frequency)
• Trie-based Prefix Search
• Greedy Ranking using Priority Queue
• String Tokenization & Normalization

---

🖥️ How to Run This Project Locally

📌 Prerequisites

• Java JDK 8 or above
• Git
• Terminal / Command Prompt / WSL

---

🔽 Clone the Repository

git clone https://github.com/adarsh-7-satyam/search-engine-java
cd search-engine-java

▶️ Compile the Project

javac *.java
java Main

✍️ Sample Input

jav ba fr

✅ Sample Output

Expanded keywords used for search:

• java
• backend
• framework

Top 5 Search Results:

1. Spring FRAMEWORK simplifies Java BACKEND development
2. Spring Boot is a Java FRAMEWORK for BACKEND systems
3. Java BACKEND developers often use Spring FRAMEWORK ...

🎯 Learning Outcomes

• Practical understanding of search engine internals

• Efficient use of DSA in real-world problems

• Experience with performance optimization

• Clean separation of search logic and presentation

📌 Future Enhancements

• Stop-word removal

• Cosine similarity ranking

• Spell correction

• REST API using Spring Boot

• Web-based UI

👨‍💻 Author

Adarsh Satyam B.Tech CSE, IIT Bhilai