Forum - Technical problems - Errors with the chatbot

Troubleshooting & FAQ Guide

If you encounter issues with the Learning-Bot plugin, the first troubleshooting step is to re-upload all original files and test again. If errors persist, refer to the solutions below.

1. How do I teach the chatbot new information?

To train the chatbot, trigger it to ask a question. While the bot selects questions randomly, it is more likely to ask one when the user provides short responses. Once the chatbot asks a question, it saves the query in an active session and adds your very next response directly to its memory database.

2. The chatbot isn't remembering conversations. What should I do?

Ensure that your server has write permissions enabled for the session/ folder inside the demo/ directory. The chatbot requires these permissions to dynamically generate temporary session text files.

3. Why is the chatbot returning extremely short answers?

• Google Translate API Credentials: If the Google Translate API is enabled, verify that your service account JSON key file is uploaded correctly to the designated google/ folder. Ensure the filename matches the sample file provided.

• Input Length & Memory Size: The bot adapts to user input. If you provide short inputs, it returns brief answers. Try submitting longer sentences and expanding the memory database with additional rows.

4. Why is speech recognition not working?

• File Permissions: Verify write permissions across all chatbot folders. The system must save a temporary raw audio file on the server prior to conversion.

• Google API Credentials: Ensure a valid Google Cloud JSON credential file is placed in the required directory.

• FFmpeg Installation: Confirm that FFmpeg is installed on your server. The engine relies on FFmpeg to re-encode raw audio into compatible formats for processing.

5. Why aren't Wikipedia descriptions being saved?

This usually occurs when no matching entries are found on Wikipedia.

• Without Google Natural Language API enabled, the bot searches Wikipedia using any words not recognized by its internal dictionary.

• Enabling the Google Natural Language API allows the system to extract precise entities and keywords from user input, significantly improving Wikipedia match accuracy.

6. Why am I getting errors when importing the database tables?

Because the dictionary contains a massive volume of entries, standard Web-based database interfaces (like phpMyAdmin) may time out due to high query limits. We recommend using a desktop database tool like HeidiSQL to execute bulk data imports reliably.

Source : learningbot

Price: $ 00.00 CAD to access learningbot extensions, forum support, chatbox support and update support