Dialogflow chatbot answers enhanced with Curation
Did you get the memo that Bots are the future of Marketing?
This quote by Riti Dass, data analyst at Mytra Jabong sums the opportunity as follows:
“Modern chatbots are, at their simplest, digital assistants that deliver specific results via a conversational interface. At their most complex, they are artificial-intelligence-powered tools that will make highly-personalized marketing scalable. They will change the face of marketing as we know it.”
The leading trailblazers in AI-based chat technology are Google Assistant, Alexa by Amazon and Siri by Apple.
According to this report by Mashable, Google Assistant abilities dominate Alexa and Siri.
DialogFlow is a Google-owned company that develops human-computer interaction technologies based on natural language conversations and is integrated with Google Assistant.
In this post, I like to show you how curation will help you design better bots faster.
Building a Dialog with Dialogflow
The main component in building dialogs in Dialogflow is called Intent.
You basically will design the flow of your conversation by defining and answering questions through this module.
Intents have two basic components, training phases and responses.
Training phases are where you input possible questions in search of an answer.
The response section then lets you build your responses and integrate with various channels like Twitter, Facebook, Slack, Twilio, and other leading channels.
In our post here we want to explore the functionality of “Google Assistant”, as this is the module we use to integrate our chat agents on our website.
Dialogflow offers us a number of response types.
Depending on the flow of your dialog you will choose one or more of these response types to lead visitors and answer their questions.
Basic Card is the most flexible of responses as it allows us to link outside the chat-bot to an external link.
By grouping responses into one link, you can help users find the response they are looking for more efficiently!
Our curation solution, KBucket lets you do this efficiently!
Flexible indexing is key to great Bot design
Designing a conversation always starts with a context.
If you are using your Bot to drive your customer service, then the context is your products and services, and if you are a political entity, then your context is your policies and positions on issues.
In either case, the context of the conversation is both defined and evolving, meaning that new information is continuously added to the fold and needs to be integrated with your response.
This is why curation is an ideal tool for building and maintaining your dialogs.
Curation with KBucket is the act of collecting, commenting and tagging content from different sources into a content hub.
As you add new information to each category, you will automatically enhance your chat response.
By curating and tagging content relevant to your project, you are in essence designing the framework for your Bot. Each category is now an intent that you can build a response to.
Let’s say you wanted to talk to people about climate change. Using your curated library, you could curate content from various sources, use climate change as the main category tag and then use related tags like climate disasters and climate solutions to focus some of your responses to more specific questions.
Your basic card in Dialogflow would look like this.
The link on the basic card will now connect you to a series of articles in your content library, supporting your response, and educating people on various aspects of climate change as you can see in the gallery below. Users simply scroll through these posts to view them.
I have created three examples to demonstrate the effectiveness of our approach.
The first Bot agent answers questions on the applications of curation for Digital Marketing, Political Campaigns, and Knowledge Management.
The next two bots are in politics. This bot is based on news curated over the past 3 years tracking the Trump presidency.
The next Bot is developed to demonstrate how political candidates can engage and inform voters by simply curating what they share on social media channels and turning it into a conversation.
Bots are a great tool for presenting information in context and helping people find answers to their questions quickly and efficiently.
Organizing your knowledge around a topic is a process that requires a flexible framework, so you can quickly and easily refine your process to achieve your goal.
The power of Dialogflow is that you can integrate your Bots onto many platforms. For each platform, you will need to modify your responses to fit the format of that particular channel.
If you like to learn more book an appointment.