Enhance Your Bot's Performance with the New Training Feature

Enhance Your Bot's Performance with the New Training Feature

Table of Contents

1. Introduction
2. The Importance of Training in Conversational AI
3. User Query Training
   1. Assigning Queries to Intents
   2. Sorting and Filtering User Queries
   3. Exploring Matched Intents
   4. Creating New Intents
4. Intent Training
   1. Aggregating User Queries
   2. Assigning Multiple User Queries to Intents
5. Intent Proposals
   1. Creating New Intents from Proposed Utterances
6. Latest Changes
   1. Tracking Modifications to Intents
   2. Adding Notes to Intent Modifications
7. Conclusion

🚀 Introducing the New Bot Friends X Training Feature

Training is an essential aspect of maintaining the performance and efficiency of chat and voice bots. With the latest release of Bot Friends X, we are thrilled to introduce our new training feature. In this article, we will delve into the importance of training in conversational AI and guide you through the various functionalities of our new training section.

💡 The Importance of Training in Conversational AI

Training lies at the core of conversational AI, enabling bots to learn and improve from user input. Unlike traditional websites or apps, conversational AI provides valuable insights into the specific information users are seeking. By analyzing user queries in plain text, the training process presents an unprecedented opportunity to enhance your bot's Knowledge Base by incorporating previously unanswered questions.

🔍 User Query Training

The user query training section is the starting point of the training process. Here, you are presented with a list of user queries that are currently not included in your bot's knowledge base. From this interface, you can assign individual queries to existing intents or create new intents altogether. User queries that are already part of the training data or existing intents are filtered out, ensuring a streamlined training experience.

When reviewing user queries, you have the option to assess the assigned intent. If an intent is misallocated, simply click on it and assign it to the correct one. For user queries that are correctly matched with intents, you can swiftly confirm the match with a single click. To optimize efficiency when dealing with large amounts of data, the sorting feature allows you to prioritize user queries based on confidence scores, starting with the lowest ones.

A notable addition to the user query training section is the ability to view the conversation context of a specific user query. By clicking on "show conversation," you can gain a better understanding of the context in which the query was made. Additionally, you now have the option to create new intents directly from the training section, empowering you to expand your bot's knowledge base and accommodate potential user inquiries.

⚡️ Intent Training

Intent training focuses on grouping and assigning user queries based on matched intents. This section provides an aggregated list of user queries that are yet to be incorporated into your bot's knowledge base. Unlike user query training, where you assign queries individually, intent training allows you to assign multiple user queries to existing intents simultaneously, greatly expediting the training process.

Within the intent training interface, you can easily identify matched intents, observe the number of utterances already assigned to each intent, and evaluate the potential user queries that can be assigned. The average confidence score of the user's query further aids in determining the most suitable assignments. This capability to efficiently allocate multiple user queries to intents significantly enhances the overall training experience.

📝 Intent Proposals

The intent proposals section presents a curated list of unassigned user queries and queries that triggered the default fallback intent. Here, you have the opportunity to create entirely new intents using the proposed sets of utterances. Unlike intent training, which focuses on refining existing intents, the intent proposals section generates suggestions for new intents based on user questions.

The proposed intents are labeled with names indicative of their content, and the number of detected user queries that Align with each proposed intent is displayed. By clicking on a proposed intent, you can explore the exact utterances that triggered it. You can select or deselect sentences, modify intent names, and even add responses. With a simple click, you can create an entirely new intent, broadening your bot's capabilities in addressing users' queries.

🔄 Latest Changes

To keep track of modifications made to intents, the latest changes table provides a comprehensive overview. This table displays the intent name, the details of modifications or updates, and the user responsible for the changes. For intent trainers, a filter option allows easy identification of newly created intents. Moreover, you now have the ability to add personal notes to intent modifications, facilitating communication and collaboration within your team.

🔖 Conclusion

In this article, we have explored the various aspects of the new training feature in Bot Friends X. User query training, intent training, intent proposals, and the ability to track and modify intents all contribute to enhancing the performance and knowledge base of your conversational AI Bot. We hope that these new features will further empower you in providing exceptional conversational experiences to your users. Stay tuned for more updates as we continue to enhance and refine these capabilities in Bot Friends X.

Highlights:

• Introducing the new Bot Friends X training feature
• Understanding the importance of training in conversational AI
• User query training: assigning queries and creating new intents
• Intent training: efficient allocation of user queries to intents
• Intent proposals: creating new intents based on proposed utterances
• Tracking and noting the latest changes to intents

FAQ:

Q: How does training benefit conversational AI? A: Training allows bots to learn and improve from user input, enhancing their knowledge base and ability to provide accurate responses.

Q: Can I assign multiple user queries to intents at once? A: Yes, the intent training section allows for the efficient assignment of multiple user queries to existing intents.

Q: Can I create new intents from proposed user queries? A: Yes, the intent proposals section provides suggested sets of utterances that can be used to create new intents.

Q: How can I track modifications made to intents? A: The latest changes table displays intent modifications, including the name of the intent, details of the changes, and the user responsible.

Q: Can I add personal notes to intent modifications? A: Yes, the latest changes feature allows you to add personal notes to intent modifications for better communication and collaboration within your team.

Resources:

• Bot Friends X