iRealtee OS Glossary

Full-Text Search

1. **Document Ingestion** — The system collects text data from property descriptions, amenities lists, developer notes, and other content fields 2. **Text Analy...

Full-Text Search

Full-text search is a text retrieval technology that searches the entire content of documents or database records to find matches based on keywords, phrases, or complex queries. Unlike simple pattern matching, full-text search analyzes text, creates searchable indexes, and ranks results by relevance.

Key Components

  • Text Indexing — The process of analyzing and storing text content in a structured format for fast retrieval
  • Tokenization — Breaking text into individual words or tokens for indexing
  • Stemming — Reducing words to their root form (e.g., "searching" becomes "search")
  • Stop Words — Common words (like "the," "and," "is") filtered out to improve search efficiency
  • Relevance Ranking — Scoring and ordering results based on how well they match the query
  • Query Parser — Interprets user search input and converts it into database-readable queries

How it Works

  1. Document Ingestion — The system collects text data from property descriptions, amenities lists, developer notes, and other content fields
  2. Text Analysis — Content is tokenized, normalized, and processed to extract meaningful terms
  3. Index Creation — Processed terms are stored in an inverted index that maps words to their document locations
  4. Query Processing — When you search, the system parses your query and matches it against the index
  5. Result Ranking — Matching documents are scored based on term frequency, field weights, and relevance factors
  6. Result Delivery — Ranked results are returned, showing the most relevant matches first

Pros & Cons

Advantages:

  • Searches across large text datasets in milliseconds
  • Handles natural language queries and partial matches
  • Supports advanced operators (AND, OR, NOT, phrase matching)
  • Returns results ranked by relevance

Disadvantages:

  • Requires additional storage for indexes
  • Index updates add overhead during data writes
  • May return false positives for ambiguous terms
  • Requires tuning for optimal results in specific domains

Real-World Application

In Philippine real estate platforms like iRealtee, full-text search enables agents to find properties using natural language. You can search for "corner lot with pool in Makati" and the system searches across property descriptions, amenities, locations, and developer notes simultaneously.

This capability helps real estate service practitioners quickly locate units with specific features from thousands of listings. Instead of browsing through filters one by one, agents type what they need and get ranked results instantly.

For property listings with detailed descriptions, full-text search matches buyer requirements against listing content — finding units that mention "near MRT," "pet-friendly," or "move-in ready" even when these aren't structured data fields.

The technology integrates with CRM systems to search client notes and communication history, helping agents recall previous conversations about specific property preferences.

References

Looking for another term? Browse the full glossary.

Back to Glossary