Sentiment analysis

Key Summary

• CREMA AI analyzes review text to quantify satisfaction levels by specific keywords.
• The system automatically categorizes shopping mall products and defines representative keywords for analysis.
• Results are displayed as positive/negative badges and detailed percentages in both review and product management menus.
• Analysis results can be used for automation settings, sorting options, and widget filters.

In this article

• Understanding CREMA AI sentiment analysis
• Checking results in Reviews & rewards
• Viewing detailed review analysis
• Analyzing sentiment in Manage by products
• Practical applications of sentiment analysis

Understanding CREMA AI sentiment analysis

CREMA AI utilizes artificial intelligence to analyze the text written by customers. It summarizes satisfaction for specific keywords into numerical data, providing a clear overview of how customers feel about your products.

• Automatic Categorization: CREMA automatically identifies the most suitable categories and representative keywords based on your shopping mall's products.
• Analysis Scope: The AI focuses on the most relevant data by analyzing reviews submitted within the last 90 days.

Checking results in Reviews & rewards

In the Review list, the AI analyzes the entire body of each review and attaches a badge to indicate the overall sentiment.

• Sentiment Badges: Reviews are tagged as either "Positive" or "Negative" based on the text content.
  
• Keyword Preview: Hover your mouse over the badge to see the primary keywords extracted by the AI. You may see:
  • No primary keywords
    
  • Only positive keywords or negative keywords
    
  • A combination of both positive and negative keywords
    
• Advanced Filtering: You can use the advanced search filters to find specific reviews based on their sentiment analysis results.
  

Viewing detailed review analysis

Clicking on a specific review in the list opens a detailed popup where you can dive deeper into the AI's findings. This section provides a comprehensive breakdown of the reviewer's sentiment and context.

• Analysis of the review body: A breakdown of sentiments found directly within that specific review text.
• Product-wide analysis results: Displays a percentage showing how other customers generally feel about this specific product.
• Reviewer's overall tendency: Shows analysis of other reviews written by the same customer. This helps you understand their typical standards for evaluation.
• Keyword-specific scoring: Provides individual scores for keywords found in the review, compared against the product's average score and the reviewer's historical average.

Analyzing sentiment in Manage by products

This menu allows you to monitor the sentiment health of your entire catalog at a glance.

• Visual Ratios: See the percentage of positive (Green) versus negative (Red) reviews for each product.
• Keyword Highlights: Identify the dominant keywords being used to describe each product.
  
• Search & Filter: Use the advanced search to filter products by:
  • Overall sentiment analysis results
  • Positive dominant keywords
  • Negative dominant keywords
    

Practical applications of sentiment analysis

The insights generated by CREMA AI can be integrated into various operational settings to improve efficiency and customer experience, such as:

• Publishing Settings: Set new reviews to "Publish automatically after AI review" if they meet positive criteria.
• Sorting Options: Allow customers to sort reviews using an "AI-recommended" option that prioritizes helpful or high-quality sentiment.
• Review Widgets: Add an "AI Keywords" search filter to your storefront widgets, helping shoppers find reviews about specific attributes (e.g., "fit," "color," "durability") quickly.

❗ IMPORTANT

• The AI analysis is based on the text body of the reviews.
• Automatic categorization relies on the product information synced from your shopping mall.
• If a review is too short or lacks descriptive keywords, the AI may not be able to generate a detailed analysis badge.