Nous Knowledge Graph
27 August 2022 | Product
For the longest time since I became CEO of Mesolitica, I wanted to understand how can we build a chatbot that could bring value to businesses.
Back in 2019, I naively thought that by understanding NLP it could be done. A ”black box” that could handle every possible business question.
This curiosity created a “radar” for me to keep on the look out for possible solutions or platform while I was learning to understand the whole AI field.
It was exciting at first, whenever new R&D come out from an AI research lab got “translated” into a module under Malaya.
Each new Language AI capability unlocked an array of business opportunity which took me down a theoretical rabbit hole of thought experiments and conceptual products.
Unlocking this “black box” led me to many possible paths.
One such capability was around the discipline of Knowledge Graph (“KG”).
Known as a semantic network, KG represents a network of real-world entities—i.e. objects, events, situations, or concepts—and illustrates the relationship between them. This information is usually stored in a graph database and visualized as a graph structure.
At it centre, the idea of KG is to help with knowledge representation and reasoning.
Conversation Context
So how does KG helps with Nous?
In general, KG helps to inject context in a particular task. Specifically for NLP, KG is used in sentence segmentation, entity and relationship extraction.
For our business chatbot Nous, KG will play a central role in capturing domain and business specific knowledge.
The approach of combining Nous NLU + Nous KG would us to deploy business chatbot with the highest business value from Day 1.
While Nous NLU (Natural Language Understanding) is tasked to capture the end user’s intent, Nous KG will give context to it with the end result of giving the correct response (and triggering the business process) — doing this correctly consistently is the objective of business chatbot.
Without Nous KG, all possible entities, relationships and business logic needs to be captured and trained for each business chatbot deployed.
During the Pre-Launch phase, we will start to develop KGs for various business types (General, Retail, Service and eCommerce)
Post-launch we would explore to extend Nous KG for specific verticals: Fashion, Healthcare, Real Estate, Finance, Telecom, Public & Government services.
Data captured for Nous would be split into two categories:
🔸 NLU for unstructured data
🔸 KG for structured data
Nous is currently in pre-launch phase. Get the offer here 👉🏻
"Nous by Mesolitica" No-code Chatbot for Business. Multimodal. Multiskill. Multilingual.
Pronounced as "nows". Meaning: practical intelligence or common sense. In Philosophy it refers to the mind, intellect, reason or intuition.
🤖🧠🇲🇾
#malaychatbot #chatbot #conversationalai #automation #businessprocessautomation #malaya #malaylanguagemodels
Khalil Nooh Co-Founder & CEO