Back to Knowledge Base

Dataset/Corpus Management

View Dataset Details and Contents

Overview

Flow ID: view-dataset-details
Category: Dataset/Corpus Management
Estimated Duration: 2-5 minutes
User Role: All Users
Complexity: Moderate

Purpose: This flow enables users to view comprehensive details about a dataset, including all the structured items (IdeaBlocks) it contains, statistics, and metadata. Users can browse through potentially thousands of items, search for specific content, and verify the quality of their dataset before using it in conversations.

Related Flows

Prerequisites

Before starting, users must have:

  • Application installed and running
  • At least one dataset uploaded or created
  • Access to Datasets page or Settings

Step-by-Step Flow

Main Path (Happy Path)

Step 1: Navigate to Datasets

  • User Action: Click "Settings" then navigate to a datasets-related tab, OR click "Datasets" in navigation if available
  • System Response: Datasets list page loads
  • UI Elements Visible:
    • List of available datasets
    • Each dataset showing:
      • Dataset name
      • Embedding model used
      • Status (Active/Inactive)
      • Item count
      • File size
      • Upload/creation date
      • Action buttons (Activate, View, Download, Delete)
  • Visual Cues:
    • Green badge for active dataset
    • Gray badge for inactive datasets

Step 2: Select Dataset to View

  • User Action: Click "View" button or click dataset name
  • System Response:
    • Page begins loading dataset details
    • Brief loading indicator
    • Dataset details page loads
  • UI Elements Visible:
    • Loading spinner briefly
    • Page transition
  • Visual Cues: Page transition animation

Step 3: Dataset Details Page Loads

  • User Action: Wait for page to load (typically 1-3 seconds, longer for very large datasets)
  • System Response: Comprehensive dataset details interface displays
  • UI Elements Visible:
    • Header Section:

      • Breadcrumb navigation (Home > Settings > Datasets OR Home > Blockify > Job > Dataset)
      • Dataset title (large, prominent)
      • Status badge (ACTIVE in green or INACTIVE in gray)
      • Action buttons: "Download" and "Activate"/"Deactivate"
      • File size indicator badge (item count + file size in MB)
    • Statistics Cards (2 cards):

      • Total Items: Large number showing total count
      • Average Content Length: Characters per item
    • Filter and Pagination Controls:

      • Search/filter input box
      • Items per page selector (10, 20, 50, 100)
      • Pagination info: "Showing 1-20 of 1,247"
      • Page navigation buttons
    • Items List:

      • 20 items displayed (default page size)
      • Each item showing:
        • Block Name (bold header)
        • Critical Question (italicized)
        • Trusted Answer (main content)
        • Source file indicator (if available)
    • Scrollable content area

  • Visual Cues:
    • Clean dashboard layout
    • Green-bordered item blocks
    • Clear information hierarchy

Step 4: Review Dataset Statistics

  • User Action: Look at statistics cards to understand dataset scope
  • System Response: Statistics displayed
  • UI Elements Visible:
    • Total Items card (e.g., "2,450 items")
    • Average Content Length card (e.g., "850 characters")
  • Visual Cues:
    • Large numbers for easy reading
    • Icons distinguish metrics
    • Cards in grid layout

Step 5: Browse Initial Items

  • User Action: Scroll through first page of items (20 items by default)
  • System Response: Items visible, scrollable
  • UI Elements Visible:
    • 20 IdeaBlock items displayed
    • Each showing complete structure:
      • Block Name: "Authentication System Overview"
      • Critical Question: "How does user authentication work?"
      • Trusted Answer: Full text content (may be several paragraphs)
    • Items clearly separated
  • Visual Cues:
    • Green borders around each item
    • Clear visual separation
    • Readable text formatting

Step 6: Optional - Filter/Search Items

  • User Action: Type search term in filter box to find specific content
  • System Response:
    • Items filter in real-time (or after brief delay)
    • Only matching items displayed
    • Pagination resets to page 1
    • Result count updates
  • UI Elements Visible:
    • Filter input with typed search term
    • Updated results: "Showing 1-20 of 45 results"
    • Only items containing search term displayed
    • "Clear filter" option (X button in search box)
  • Visual Cues:
    • Instant filtering (smooth)
    • Result count changes
    • Clear indication results are filtered

Step 7: Navigate Between Pages

  • User Action: Click page numbers or next/previous buttons to view more items
  • System Response:
    • Page changes
    • New set of items loads
    • Pagination info updates
  • UI Elements Visible:
    • Page navigation controls:
      • "Previous" button (disabled on page 1)
      • Page numbers (1, 2, 3, ... with current highlighted)
      • "Next" button (disabled on last page)
    • Results info: "Showing 21-40 of 1,247"
    • New items display
  • Visual Cues:
    • Active page number highlighted (blue)
    • Smooth content transition
    • Page number updates

Step 8: Optional - Adjust Items Per Page

  • User Action: Click items per page selector and choose different value (e.g., change from 20 to 50)
  • System Response:
    • Page reloads with new page size
    • More items displayed
    • Pagination updates
  • UI Elements Visible:
    • Dropdown showing: 10, 20, 50, 100 options
    • 50 items now displayed (or selected amount)
    • Fewer total pages
    • Updated pagination: "Showing 1-50 of 1,247"
  • Visual Cues:
    • More items visible at once
    • Longer scroll area

Step 9: Optional - Download Dataset

  • User Action: Click "Download" button in header
  • System Response:
    • Dataset file generates
    • Download begins
    • File saves to computer
  • UI Elements Visible:
    • Download button in header
    • Browser download notification
    • Progress indicator
  • Visual Cues: Download in progress
  • Note: See export-dataset.md for detailed download flow

Step 10: Optional - Activate Dataset

  • User Action: Click "Activate" button if dataset is currently inactive
  • System Response:
    • Dataset activates
    • Status badge changes to "ACTIVE" in green
    • Button changes to "Deactivate"
  • UI Elements Visible:
    • Status badge updates
    • Button text changes
  • Visual Cues:
    • Green active indicator
    • Button state change
  • Note: See corpus-activation.md for detailed activation flow

Step 11: Return to Previous Page

  • User Action: Click breadcrumb link or back button
  • System Response: Navigate back to datasets list or previous page
  • UI Elements Visible: Previous page content
  • Visual Cues: Page transition

Final Step: Dataset Examined

  • Success Indicator:
    • Viewed dataset contents and statistics
    • Understood what information dataset contains
    • Verified data quality
    • Can search and navigate dataset effectively
  • System State Change:
    • User has knowledge of dataset contents
    • Can make informed decision about using dataset
    • May have activated dataset if desired
  • Next Possible Actions:
    • Activate dataset for use in conversations
    • Download dataset for backup or sharing
    • Return to chat and query this dataset
    • Upload additional datasets
    • Create new blockify job to add to dataset

Alternative Paths & Strategies

Strategy A: Quick Verification Check

When to use: Just need to confirm dataset has content

Steps:

  1. Navigate to dataset details
  2. Glance at total items count
  3. Scan first few items to verify structure
  4. Return to previous page
  5. Total time: 30 seconds

Strategy B: Comprehensive Quality Review

When to use: Need to thoroughly verify dataset quality

Steps:

  1. Open dataset details
  2. Review statistics carefully
  3. Browse multiple pages of items
  4. Search for known topics to verify coverage
  5. Check for duplicates or errors
  6. Test various search terms
  7. Document findings
  8. Total time: 10-20 minutes

Strategy C: Search-Focused Review

When to use: Looking for specific information

Steps:

  1. Open dataset details
  2. Immediately enter search term
  3. Browse filtered results
  4. Try different search terms
  5. Verify desired information exists
  6. Note which items contain needed info

Strategy D: Access from Job Details

When to use: Viewing dataset immediately after job completion

Steps:

  1. From job details page, click "View Dataset" button
  2. Automatically navigates to dataset details
  3. Breadcrumb shows path: Job > Dataset
  4. Can return to job via breadcrumb
  5. Convenient workflow from creation to verification

Error States & Recovery

Error 1: Dataset Not Found

Cause: Dataset deleted or database error
User Experience:

  • Error message: "Dataset not found" or 404 error
  • Empty or error page
  • Cannot view contents

Recovery Steps:

  1. Return to datasets list
  2. Verify dataset exists
  3. Dataset may have been deleted
  4. If exists in list, try viewing again
  5. If persists, indicates data corruption

Error 2: Cannot Load Items

Cause: File error or corrupted dataset
User Experience:

  • Page loads but items area shows error
  • Message: "Failed to load items" or "Error reading dataset"
  • Empty items list

Recovery Steps:

  1. Refresh page
  2. Try again
  3. Check if dataset file exists (file size shown in header)
  4. If file missing or corrupted, may need to re-create dataset
  5. Check source job if dataset created via blockify

Error 3: Search Returns No Results

Cause: Search term doesn't match any content
User Experience:

  • Empty results area
  • Message: "No items match your search"
  • Result count: "Showing 0 items"

Recovery Steps:

  1. Try different search terms
  2. Check spelling
  3. Use more general terms
  4. Clear filter to see all items again
  5. Content may genuinely not exist in dataset

QA Note: Not an error - expected behavior when search term doesn't match. Documented as user experience scenario.

Error 4: Pagination Broken

Cause: Display or calculation error
User Experience:

  • Cannot navigate to next page
  • Page buttons don't work
  • Stuck on one page

Recovery Steps:

  1. Refresh page
  2. Try different page size
  3. Clear filter if applied
  4. Navigate using URL if direct page access supported

Version History

Date Version Author Changes
2025-10-04 1.1 Iternal Technologies Initial comprehensive documentation

Notes

Important Considerations:

  • Large datasets may take several seconds to load initially
  • Pagination helps manage thousands of items efficiently
  • Search is real-time with brief delay to avoid excessive filtering
  • Each IdeaBlock shows: Block Name, Critical Question, and Trusted Answer
  • Items are in creation order unless filtered
  • Cannot edit items directly; must re-create dataset to modify

Best Practices:

  • Use search feature to verify dataset contains expected information
  • Review first page carefully to check data quality
  • Browse multiple pages to get sense of full dataset
  • Use larger page sizes (50-100) for faster browsing if comfortable
  • Take notes on coverage gaps for future dataset improvements
  • Verify critical information exists before relying on dataset

Common User Questions:

  • "Can I edit items?" - No, dataset is read-only; must re-create to modify
  • "Why is it slow to load?" - Large datasets (thousands of items) take time to load and display
  • "Can I delete individual items?" - No, cannot edit; must re-create dataset without unwanted items
  • "How do I search multiple terms?" - Type multiple words; search matches items containing all words
  • "What if I can't find something?" - May not be in dataset; verify in source documents or add to dataset

Trigger

What initiates this flow:

  • User manually initiates

Specific trigger: User wants to examine dataset contents, typically because:

  • They want to verify what information is in the dataset
  • They need to search for specific topics or items
  • They're validating dataset quality after creation
  • They want to understand dataset coverage before activating it
  • They need to review items for accuracy or completeness

User Intent Analysis

Primary Intent

Browse and examine the complete contents of a dataset to understand what information it contains and verify its quality and usefulness for AI conversations.

Secondary Intents

  • Verify dataset creation succeeded
  • Search for specific information within dataset
  • Assess data quality and accuracy
  • Understand dataset scope and coverage
  • Identify gaps or issues in content
  • Determine if dataset is suitable for intended use

Subintents

  • Navigate through potentially large datasets efficiently
  • Find specific items or topics quickly
  • Understand item structure and format
  • Verify embeddings and search functionality
Free Trial

Download Blockify for your PC

Experience our 100% Local and Secure AI-powered chat application on your Windows PC

✓ 100% Local and Secure ✓ Windows 10/11 Support ✓ Requires GPU or Intel Ultra CPU
Start AirgapAI Free Trial
Free Trial

Try Blockify via API or Run it Yourself

Run a full powered version of Blockify via API or on your own AI Server, requires Intel Xeon or Intel/NVIDIA/AMD GPUs

✓ Cloud API or 100% Local ✓ Fine Tuned LLMs ✓ Immediate Value
Start Blockify API Free Trial