Research Center for AI and Personalization

Information Retrieval (IR) Evaluation Navigator IREN

USA flag English
USA flag English
Russian flag Русский
Chile flag Español
Brazil flag Português

IREN

Advanced AI-powered search evaluation platform

Load Session

Open Evaluation

Continue your previous evaluation work by loading a saved session file

Start New Evaluation

Start New Evaluation

Begin a new search evaluation project from scratch

Learn System

Learn System

Explore documentation and learn how to use IREN effectively

Sampling

Manage search query datasets with distribution analysis and file operations.

Sample Management

Awaiting file selection...

Manual Entry

Search Phrase Sample

0 search phrases
Search Phrase Frequency Actions

Classification

Enrich the sample by assigning each query an Information-Need category, grouping them into Semantic Clusters, labelling User Intents, and calculating a Conversion-Potential score.

Advanced Mode

Classification Method

0.7

Classification Rules

BERT Model Configuration

Clustering Configuration

User Intents Determination

This method uses predefined keywords to classify queries as Transactional, Informational, or Navigational.

Conversion Potential Scoring

Each factor is multiplied by its weight and then added together. The theoretical range of the Conversion Score with the current weights is calculating… Adjust weights or keyword lists to suit your data.

Labeling

Labeling is a critical alignment stage. For every search phrase we define the authoritative product or category the search engine should surface, record the live inventory quantity and attach the target URL. This step normally requires close collaboration between merchandisers, search-relevance engineers and business stakeholders so that the expected results are explicit and agreed by all parties.

Labeling Configuration

Labeling Progress

0 / 0 completed

Engine Evaluation

Comparative search engine performance assessment with advanced analytics

Search Engine Configuration

MANDATORY

Engine A

URL Encode
ACTIVE
OPTIONAL

Engine B

URL Encode
INACTIVE
OPTIONAL

Engine C

URL Encode
INACTIVE

Evaluation Configuration

Batch Screenshot Capture

Capture search results pages in bulk across engines for the evaluation sample.

Evaluation Results

Comparative search engine performance assessment with advanced analytics

Engine Comparative Results

Win Analysis ?

Performance Visualization

Evaluation Results

Query Engine Total in SERP Relevant in SERP Total in Inventory Precision Recall F1-Score

Findings & Recommendations

Analytics

Comprehensive analytics dashboard (placeholders)

Overview & Summary

Click to open high-level metrics and quick insights.

User Intents

Placeholder for intent distribution and trends.

Semantic Clusters

Placeholder for cluster visualizations and summaries.

Conversion Optimization

Placeholder for conversion opportunity analysis.

Topic Analysis

Placeholder for emerging and declining topics.

Long Tail Analysis

Click to open long-tail metrics and charts.

Hierarchical Structure

Placeholder for category hierarchy insights.

Query Explorer

Placeholder for interactive query drill-down.

Long Tail Analysis

Analyze the distribution of search query volume to identify head and long-tail patterns.

Head Value (Up to 80% Volume)

0
Sum of (Volume * Score) for 0 Searches

Cumulative Volume Threshold

%
Search volume 0 at threshold point

Tail Value (After 80% Volume)

0
Sum of (Volume * Score) for 0 Searches

Tail Value Share

0%
Tail's % of Total Weighted Value

Automated Findings & Recommendations