Chatbots for Monetizing Expertise

A New Era for SMEs to Monetize Their Knowledge!

Every person can be some sort of subject-matter-expert (SME) regardless of the complexity of the subject. Knowledge acquired over the years (thru education, work, etc.) is unfortunately bottle-necked by the limited communication bandwidth of our biological framework. We can only tutor one or two person at a time with the highest level of participation. We may give speeches to a group of people with less level of interaction. Finally, writing essays, books, blogs, videos, or social media gives us broader audience, but with minimal interaction. As a result, expertise is highly limited in its interactive exposure.

Chatbots can scale up interactive exposure of expertise 1000s of times more than peer to peer alternative, which is like one-on-one tutoring 1000s of users, simultaneously.

Editorial Platform to Create Expert Chatbots
Building SME chatbots must be refined to content curation and persona creation without any coding or AI training requirements. Otherwise, SME chatbots will become very expensive and lenghtly to develop that will jeopardize their ROI. The challenge is to turn content into a sense-making chatbot with dialogue and short-term memory skills. For this, various machine learning platforms may be used. exClone’s platform is an example where SME chatbots can be build fast using samples loaded in users’ accounts.

An example to interactive exposure of a subject matter expertise is shown below for tax advice for small business owners derived from the IRS Publication 334.

Expert Chatbot is a Conversational Flow Chart
Some people do not fully appreciate the difference between a chat interaction (dialogue) and flat document search. This is also wrongfully encouraged by products like Siri and Alexa where the conversation is a single-step question/answer. Human dialogue, which is the ultimate form of learning/teaching, is beyond a single step communication. It requires rather complicated short-term memory ability, because the user has many steps to describe his/her problem. Here is a very simple example, a flow chart for handling your taxes involving health insurance.

There are five different actions you can take according to the flow chart about how to handle your taxes involving health insurance. It is instantly obvious that you could not state your problem in one Google query. Flat document search will not work by any means. Rather, you either need a diagram like this, or an expert who walks you through a similar work flow. A chatbot embedded with this expertise could be just as useful especially if more information is tucked under each block for impromptu questions.

Most expert solutions involve multiple steps through a flow of information exchange that cannot be replicated by any Google Query.

Flow charts define a dialogue template in exClone platform some of which are pre-defined whereas others are customized.
Knowledge-based Machine Learning Challenge
The scientific challenge in creating expert chatbots is the ability to convert flat documents, (like the IRS Publication 334), and flow charts (like the one shown above) into proper dialogue flow in the form of a chatbot. Because the knowledge is already available, the machine learning challenge is to automate this process without any coding effort. Obviously, data-driven machine learning methods (such as deep learning) do not apply here since knowledge is already available. This crucial different was explained in one of my earlier articles titled “Can Machine Learning Use Knowledge instead of Data? Deep Cloning vs Deep Learning“.


This article is brought to you by exClone, a chatbot technology provider.

Chat with Vera about exClone.

Try free (no cc required) of our Cloning Platform via Linkedin access.

Join CHATBOTS group in linkedin.

You can follow exClone in Facebook, and in LinkedIn.