Reviu - Game Review Explorer

Reviu is a professional video game review search system that allows users to explore and discover game reviews through an intuitive interface. The system features advanced search capabilities, including semantic search, category filtering, and automatic topic clustering of results.

Features

• Advanced Search Options

  • Traditional keyword-based search with query boosting
  • Semantic search using sentence embeddings
  • Category-based filtering (Controls, Multiplayer, Story, Technical, Relaxing)
  • Minimum score filtering

• Dynamic Result Clustering

  • Automatic grouping of search results into relevant topics
  • Interactive cluster exploration
  • Detailed review viewing with similar game recommendations

• Latest Reviews

  • Showcase of recent game reviews
  • Random selection for variety

Technology Stack

• Backend

  • Flask (Python web framework)
  • Apache Solr (Search engine)
  • sentence-transformers (Semantic search capabilities)
  • scikit-learn (Results clustering)

• Frontend

  • HTML5/CSS3
  • Tailwind CSS
  • JavaScript (Vanilla)

Installation & Setup

1. Prerequisites

   # Install Docker and Docker Compose
   # Install Python 3.8+

2. Clone the Repository

   git clone [repository-url]
   cd reviu

3. Environment Setup

   python -m venv venv
   source venv/bin/activate  # On Windows: .\venv\Scripts\activate
   pip install -r requirements.txt

4. Start Solr

   make up
   make setup-cores
   make setup-semantic

5. Index Data

   make index-data
   make process-semantic

6. Run the Application

   python app.py

The application will be available at http://localhost:5000

Query Examples

The system supports various pre-defined queries for different gaming aspects:

// Controls & Gameplay Query
{
  "params": {
    "q": "Content:(controls good tight precise responsive fluid excellent smooth)",
    "rows": 30,
    "fl": "id,Title,Score,Subtitle,Content",
    "sort": "score desc"
  }
}

// Multiplayer Experience Query
{
  "params": {
    "q": "Content:(multiplayer online coop cooperative competitive pvp gameplay fun engaging players teams)",
    "rows": 30,
    "fl": "id,Title,Score,Subtitle,Content",
    "sort": "score desc"
  }
}

Interface Screenshots

Main search interface with category filters and search options

Example of clustered search results

Detailed review view

Development

• Adding New Queries

  • Place query JSON files in the queries directory
  • Follow the existing query format
  • Update the Makefile with new query targets

• Modifying Search Behavior

  • Edit construct_solr_query() in app.py
  • Update Solr schema configurations in docker/solr/conf/

• Customizing Result Clustering

  • Modify the cluster_results() function in app.py
  • Adjust category keywords and scoring logic

Acknowledgments

• IGN for the game reviews dataset