Bot Creation Simplified with Amazon Lex V2

Overview

A chatbot is a software that simulates human conversations with users via text chat. Its main task is to help users by providing answers to their questions. If the chatbot also includes voice features, then it is called a voice bot. AWS provides a service called Amazon Lex which can help you to create voice bots along with Amazon Polly which gives the speaking capability to the bot. Basically, there are two versions of Amazon Lex. Now we can look into what exactly is Amazon Lex V2.

Introduction to Amazon Lex V2

Amazon Lex V2 is an AWS service that is used to build conversational interfaces using both text and voice.  Amazon Lex V2 can help any developer to build bots quickly. To create a bot, you just need to give basic conversational flow. Using the Amazon Lex V2 console you can build, test, and then deploy the bot to your applications. Amazon Lex provides integrations with AWS Lambda, Amazon Connect, and Amazon Kendra. This doesn’t need prior Deep Learning experience. Deep functionality and flexibility of natural language understanding, and automatic speech are provided by Amazon Lex V2.

Amazon Lex Terminologies

1. Bot: A bot can perform automated tasks like booking a flight ticket, booking a hotel, etc. This can understand the voice or text input provided by the user and converse using natural language.
2. Language: Amazon Lex V2 supports multiple languages in which the bot can converse. Every language is independent. For example, the locales of English that are supported by Amazon Lex V2 are English (UK), English (India), English (US), English (South Africa), and English (Australia).
3. Intent: The intent says what the user wants to perform. One bot can have one or more intents in it. The information to be provided while creating an intent are:

• Intent name: Name of an Intent.
• Sample Utterance: What the user might say to trigger an intent.
• How to fulfill intent: How users want to fulfill intent after providing required information.
• Note: Amazon Lex V2 has built-in intents and contains the fallback intent by default, fallback intent is triggered when the bot is not able to detect the user’s intent.

4. Slot: A intent has slots or parameters that prompt to take user information. One intent can have one or more slots in it. The user should provide all the slot values before Amazon Lex V2 fulfill the intent.
5. Slot Type: Every slot has a type. One can use inbuilt slot types which are provided by AWS or create your own custom slots. For example, there is an inbuilt slot type called AMAZON. PhoneNumber which can be used to take phone numbers as user input.
6. Version: A version is a numbered snapshot of your work that you can publish for use in different parts of your workflow. Once you create a version it will stay the same while you continue to work until the new version is created.
7. Alias: Alias is a pointer to the version. Using an alias, we can update the version of the bot that is being used. For example, if an application is using version 1 and you also have created version 2, now you want to use version 2, then you can simply point the alias to version 2 because the application used is alias not version.

Benefits of using Amazon Lex V2

• Simplicity: Amazon Lex V2 will help create a bot simply by providing a few phrases and the required slots and intents in a few minutes. It can build a complete natural language model that interacts with the user with both voice and text.
• Seamless Deployment and Scaling: Amazon Lex V2 can build, test, and publish the bot directly through the console. This can also integrate with your mobile applications and web apps. This can scale automatically and no need to worry about hardware provisioning.
• Built-in integration with the AWS platform– Amazon Lex V2 can work with other AWS services, such as AWS Lambda and Amazon CloudWatch. You can take the benefits of the AWS platform for security, monitoring, storage, and mobile app development.
• Cost-effectiveness–The pay-as-you-go pricing and the low cost per request make the service a cost-effective way to build conversational interfaces. With Amazon Lex V2, there are no upfront costs or minimum fees. You are charged only for the text or speech requests that are made.

Limitations of Amazon Lex

• Amazon Lex V2 has a difficulty in identifying Indian names through voice.
• Amazon Lex V2 also has a difficulty in identifying the phone numbers provided by users through voice. For example, the phone number cannot be spelled as double nine, triple six, or ninety-eight, fifty-six.
• The email slot of Amazon Lex V2 expects the user to spell ‘@’ as ‘at’.

Conclusion

Amazon Lex V2 is an easy-to-use service using which we cant create a bot with just a few clicks. This won’t have any upfront cost as it follows a pay-as-you-go model. This supports multiple languages, we can build the same bot in multiple languages. This also has an import and export feature using which one can import the bot created in another AWS account or export the bot to another AWS account.

About CloudThat