Home AI Chatbot

AI Chatbot

By Abdurrahman
3 articles

AI Chatbot Quick Start Guide

What is AI Chatbot? AI Chatbot is an intelligent assistant that helps you explore and analyze CKAN datasets using natural language. Simply type questions in plain English and get insights, visualizations, and data analysis. Getting Started in 3 Steps Step 1: Access AI Chatbot Navigate to the AI Chatbot interface from your CKAN portal. You'll see a welcome screen with example queries. Step 2: Ask Your First Question Type a question in the input box at the bottom. Try: - "What datasets are available?" - "Show me population data" - "Create a chart of unemployment rates" Step 3: Understand the Response The AI will provide: - Answer: Direct response to your question - Datasets Found: Relevant data sources - Visualizations: Charts if requested - Thinking Process: How the AI reached its conclusion (expandable) Interface Overview Header Controls - ? Thinking Toggle: Show/hide AI's reasoning process - ? Export: Download your conversation - ?️ Clear: Start a new conversation - ❓ Help: Access this guide Message Area - Your messages: Appear on the right (blue) - AI responses: Appear on the left with robot icon - Interactive elements: Click to expand datasets, view charts Input Area - Type questions (1000 character limit) - Press Enter or click Send - Character counter shows remaining space Key Features 1. Natural Language Understanding Ask questions naturally: - "What's the unemployment rate in California?" - "Compare education levels between states" - "Show trends over the last 5 years" 2. Smart Dataset Discovery The AI automatically: - Searches all available datasets - Shows relevance scores - Provides direct links to data 3. Instant Visualizations Request charts by asking: - "Create a bar chart of..." - "Show a line graph of..." - "Visualize this data" 4. Context Awareness The AI remembers your conversation: You: "Show unemployment data" AI: [provides response] You: "Now add education levels" ← AI knows you mean to the unemployment data Basic Commands What You WantHow to AskFind data"What datasets are available about [topic]?"View data"Show me data from [dataset name]"Analyze"What's the average/total/trend in [data]?"Compare"Compare [A] and [B]"Visualize"Create a chart showing [data]"Summarize"Summarize the key findings" Tips for Better Results Be Specific ❌ "Show data" ✅ "Show population data for Texas cities over 100,000 people" Include Timeframes ❌ "What's the rate?" ✅ "What's the unemployment rate for 2023?" Request Output Format - "Show this as a table" - "Create a bar chart" - "List the top 10" Next Steps 1. Try the example queries on the welcome screen 2. Explore your specific area of interest 3. Experiment with different visualization types 4. Check out the Examples Guide for more ideas Need Help? - This is a beta feature - verify critical information - Check the FAQ for common questions

Last updated on Feb 24, 2026

AI Chatbot Examples & Use Cases

Common Query Examples Finding Datasets Basic Search "What datasets do you have?" "Show me all health data" "Find education datasets" Specific Criteria "Show datasets with more than 10,000 records" "What data was updated this week?" "Find all datasets from California" ? Data Analysis Basic Statistics "What's the average income by state?" "Show me the total population of Texas" "Calculate the median home price" "What's the highest unemployment rate?" Trend Analysis "Show unemployment trends over the last 5 years" "How has population changed since 2010?" Creating Visualizations Simple Charts "Create a bar chart of population by state" "Show a pie chart of budget allocation" "Plot crime rates by city" Complex Visualizations "Create a bar chart of the top 10 cities by population, sorted descending" "Show a multi-line graph comparing unemployment across regions" "Create a stacked bar chart of energy sources by year" "Make a scatter plot of income vs education level with trend line" Comparisons Simple Comparisons "Compare California and Texas population" "Show the difference between 2022 and 2023 sales" "Which state has the highest GDP?" "Compare urban vs rural education rates" Multi-factor Analysis "Compare states by population, GDP, and unemployment rate" "Show how education levels correlate with income across regions" "Analyze the relationship between crime rates and population density"  Step-by-Step Examples Example 1: Regional Economic Analysis Step 1: Start broad You: "Show me economic data for California" Step 2: Focus on specifics You: "Focus on unemployment and GDP for the last 5 years" Step 3: Add comparison You: "Compare this with Texas and New York" Step 4: Visualize You: "Create a line chart showing trends for all three states" Step 5: Analyze You: "What patterns do you see? Which state is performing best?" Example 2: Education Research Step 1: Find relevant data You: "What education datasets are available?" Step 2: Explore the data You: "Show me high school graduation rates by state" Step 3: Deep dive You: "Which factors correlate with graduation rates?" Step 4: Time analysis You: "How have graduation rates changed over the past decade?" Step 5: Generate insights You: "Summarize the key findings and create a report" Query Templates For Analysis - "Analyze [dataset] for [specific metrics]" - "What trends exist in [data] from [start date] to [end date]?" - "Calculate [statistic] for [dataset] grouped by [category]" For Comparison - "Compare [metric] between [group A] and [group B]" - "Show the difference in [metric] before and after [date/event]" - "Rank [entities] by [metric]" For Visualization - "Create a [chart type] showing [data] by [category]" - "Visualize the relationship between [variable 1] and [variable 2]" - "Generate a dashboard with [list of metrics]" For Discovery - "What datasets contain information about [topic]?" - "Find all [data type] from [source/date/location]" - "Show me what data is available for [specific need]" Pro Tips 1. Build on Previous Queries: The AI remembers context 2. Combine Multiple Operations: Ask for analysis + visualization 3. Iterate and Refine: Start simple, then add complexity 4. Specify Output Format: Tell the AI how you want results presented 5. Use Natural Language: Don't worry about technical syntax

Last updated on Feb 24, 2026

AI Chatbot FAQ & Troubleshooting

Frequently Asked Questions General Questions Q: What is AI Chatbot and how does it work? A: AI Chatbot is an AI-powered assistant that understands natural language queries about your CKAN datasets. It searches, analyzes, and visualizes data based on your questions. Q: How accurate is the AI? A: The AI is highly accurate but currently in beta. Always verify critical information independently, especially for important decisions. Q: Can I save my conversations? A: Yes! Use the Export button (?) in the header to download your conversation as a markdown file. Q: Is there a limit to how many questions I can ask? A: This depends on your account type. Most users have generous limits for normal usage. Q: Does the AI remember previous questions in a session? A: Yes, the AI maintains context throughout your conversation. You can reference previous responses. Data & Privacy Q: Can the AI access my private datasets? A: The AI can only access datasets you have permission to view in the CKAN portal. Q: Is my conversation data stored? A: Conversations are stored temporarily for your session. Check your organization's data policy for details. Q: Can others see my queries? A: No, your conversations are private to your account. Q: What happens to my data when I clear the conversation? A: Clearing removes the conversation from your current session. Exported conversations remain in your downloads. Features & Capabilities Q: What types of charts can the AI create? A: Bar charts, line graphs, pie charts, scatter plots, heatmaps, and more. Just specify what you want! Q: Can the AI combine data from multiple datasets? A: Yes, the AI can analyze and compare data across multiple datasets when relevant. Q: What file formats are supported? A: CSV, JSON, Excel (XLSX/XLS), and other formats supported by your CKAN instance. Q: Can I modify the charts after they're created? A: You can request modifications by asking follow-up questions. For manual editing, download the chart. Q: Does the AI work with real-time data? A: The AI works with the most recent data available in your CKAN instance. Real-time depends on how often datasets are updated. Troubleshooting Guide Common Issues and Solutions ? "AI Not Understanding My Query" Symptoms: - Wrong data returned - Misinterpreted question - Irrelevant response Solutions: 1. Rephrase using simpler language 2. Break complex questions into parts 3. Use specific dataset names if known 4. Add more context Example Fix: Instead of: "Show the thing from last year" Try: "Show unemployment data from 2023" ? "No Datasets Found" Symptoms: - "I couldn't find any relevant datasets" - Empty results Solutions: 1. Check spelling of technical terms 2. Try alternative keywords 3. Broaden your search terms 4. Verify the data exists in CKAN Example Fix: Instead of: "labour statistics" Try: "employment data" or "job statistics" or "workforce data" ? "Charts Not Displaying" Symptoms: - Blank chart area - "Chart not available" message - Loading indefinitely Solutions: 1. Refresh the page (F5) 2. Check browser console for errors (F12) 3. Try a simpler chart type 4. Reduce data points (ask for "top 10" instead of all) 5. Clear browser cache ? "Slow Response Times" Symptoms: - Long wait for responses - Timeout errors - "Still thinking..." for extended periods Solutions: 1. Simplify your query 2. Request smaller date ranges 3. Ask for specific datasets rather than broad searches 4. Avoid analyzing very large datasets (>100k records) Example Fix: Instead of: "Analyze all historical data" Try: "Analyze data from the last 3 years" ? "Login Required Message" Symptoms: - "Please log in to use the AI assistant" - Session expired warnings Solutions: 1. Click the login button 2. Check if your session expired (re-login) 3. Verify your account has AI access permissions 4. Clear cookies and login again ? "Incorrect Calculations or Data" Symptoms: - Wrong numbers in response - Calculations don't match source - Inconsistent results Solutions: 1. Ask the AI to recalculate 2. Request to see the source data 3. Check the thinking process for errors 4. Verify dataset version/date 5. Report persistent issues ? "Context Lost Mid-Conversation" Symptoms: - AI doesn't remember previous questions - Have to repeat information - "I don't have that information" for follow-ups Solutions: 1. Reference the specific previous response 2. Summarize context in your question 3. Check if session timed out 4. Avoid very long conversations (export and start fresh) Example Fix: Instead of: "Add population to that" Try: "Add population data to the unemployment chart you just created" Tips & Tricks Getting Better Results 1. Use Specific Dates 2. ❌ "recent data" 3. ✅ "data from January 2023 to December 2023" 4. Name Your Outputs 5. ❌ "make a chart" 6. ✅ "create a bar chart" 7. Specify Groupings 8. ❌ "show by location" 9. ✅ "group by state" or "group by city" 10. Set Limits 11. ❌ "show all cities" 12. ✅ "show top 20 cities by population" Understanding the Thinking Process The "Thinking Process" shows: - Search queries used to find data - Datasets examined - Analysis steps performed - Decisions made by the AI Click to expand and understand how the AI interpreted your question. Error Messages Explained Error MessageWhat It MeansWhat to Do"Dataset not found"The specific dataset doesn't existCheck dataset name/ID"Insufficient data"Not enough data for analysisBroaden your query"Query too complex"Too many operations requestedBreak into smaller queries"Rate limit exceeded"Too many requestsWait a few minutes"Invalid date range"Dates don't make senseCheck date format Still Need Help? Contact Support If issues persist after trying these solutions: 1. Export your conversation for reference 2. Note the exact error message 3. Contact your CKAN administrator 4. Include browser and version info Report a Bug Help improve AI Chatbot by reporting: - Consistent errors - Incorrect calculations - UI/display issues - Feature requests

Last updated on Feb 24, 2026