AIML Introduction

AIML (Artificial Intelligence Markup Language) is a description language used in the development of natural language software agents like chatbots and virtual assistants. AIML was developed by Richard Wallace from 1995 to 2000, and it is based on XML (eXtensible Markup Language). AIML provides automated responses to users’ questions because it is rule-based. The scripting language represents a data-driven programming paradigm. While writing the script for a conversational chatbot, all the conditions (the user’s query) and actions (the chatbot’s response) must be included.

Features of AIML

• data-driven: The user’s queries are predefined as conditions and mapped with particular actions for generating automated responses for users.
• Rule-based Scripting language: If the conditions get satisfied, it performs an action based on the rule; if the condition does not get satisfied, then it stops.
• XML-based: AIML is easy to learn for programmers who know HTML or XML because it has a similar syntax to HTML and XML. It enables integration with XML editors using XML syntax.
• Flexibility: Customized tags and functions can be created for efficient code writing in development.
• Random Actions: Based on a single condition, multiple actions can be defined.

AIML Tags and their syntax

There are multiple elements and tags in AIML and we are going to see some basic elements with their syntax.

1. <aiml>: Inside this tag, all tags are defined. It constructs the starting and ending of an AIML document.

<aiml> <!-- all the tags that contain conditions and actions--> </aiml>

2. <category>: inside each category tag defines conditions and actions. category tags can be multiple inside the <aiml> tag.

<aiml>
  <category>
<!--condition and action are defined-->
  <category>
</aiml>

3. <pattern>: The condition is written inside the pattern tag.

<aiml>
  <category>
<!--condition and action are defined-->
  <category>
</aiml>

4. <template>: The action is written inside the template tag. Actions can be single or multiple, so the template tag is defined multiple times inside the category tag.

<aiml>
  <category>
    <pattern> <!--condition is defined--> </pattern>
    <template> <!--Action is defined for condition--> </template>
  <category>
</aiml>

AIML Examples

AIML Script is an XML Document, that contains elements defined using AIML schema. It must contain only one <aiml> element. The conditions and actions are defined within <category> element. For condition, <pattern> tag is used and for action, <template> tag is used.

Example 1: In this example, the condition is Hello and for that condition, the action is Hi, how can I help you? you can include multiple conditions and actions using multiple <category> elements.

XML

<?xml version="1.0" encoding="UTF-8"?>
<aiml>
    <category>
        <pattern>Hello</pattern>
        <template>Hi, how can I help you</template>
    </category>
</aiml>

Output: 

Example 2: In this example, we just continue the same process with another example by taking the chat process further.

XML

<?xml version="1.0" encoding="UTF-8"?>
<aiml>
    <category>
        <pattern>Hello, I need some help</pattern>
        <template>Hi, how can I help you</template>
    </category>

    <category>
        <pattern>How to learn JavaScript</pattern>
        <template>You can learn JavaScript from its official documentation</template>
    </category>
</aiml>

Output: 

Applications of AIML

• FAQs: The FAQs (frequently asked questions) can be mapped with their answers using the data-driven method of AIML.
• Virtual Assistant: Service-provider companies can use virtual assistance for their customers to resolve their queries. It can also troubleshoot problems.
• Personal Assistant: The conversational agents can be used as personal assistants for humans to schedule appointments and daily tasks.
• Education: The students can use AIML chatbots to learn the chapters from subjects using AIML-based chatbots.

Advantages of AIML

• Interacts with as a person: human-like conversational skills can be developed using AIML for the conversational agents.
• Natural language processing: It can process human languages to provide the desired response to the user.
• Developer-friendly: very easy to learn for the developer who knows XML and HTML.
• Improves Efficiency: Repetitive tasks and FAQs (frequently asked questions) with their responses can be automated.
• Requires low resources: the deployment of a chatbot requires low consumption of memory and processing power.

Disadvantages of AIML

• Analytical Features: AIML does not provide advanced natural language processing features like sentimental analysis.
• Learning: AIML-based chatbots rely on data provided by the developer; if new queries occur, AIML chatbots cannot learn new things to respond to those queries for the user.
• Dependent on data: using AIML, the chatbots can be developed on the basis of the provided data only. It cannot automatically update according to new changes in the world.

Conclusion

The conclusion of the article is that AIML is a developer-friendly scripting language used for creating chatbots and assistance based on XML document syntax. Every NLP-based agent developer needs to learn this AIML description language. Simple chatbots can be easily created, but for complex tasks, AIML has some limitations. Still,  AIML is a very good choice for creating conversational basic chatbots.