Most users have become accustomed to playing a guessing game in matching a search query to a relevant result. But when the stakes are as high as matching problems to solutions for business challenges on a national scale, there simply can be no room for error.

IMDA’s Open Innovation Platform crowdsources tech

A government crowdsourcing initiative does exactly this—match real business challenges of tech-buying organisations to a pool of solution providers who can develop appropriate tech solutions to the problems. Called the Open Innovation Platform (OIP), the online ecosystem gives both stakeholders better visibility to real business challenges and tech capabilities, so they can collaboratively discover innovation and stimulate business opportunities.

Major updates coming soon

Since its launch in 2018 by the Infocomm Media Development Authority (IMDA), the platform has matchmade over 190 challenges and 60 solutions, but is now primed for an update with a $50 million government investment to further build on its successes. In particular, the OIP’s search functionality will be receiving a major update to further propel rapid innovation discovery.

OIP’s new Discovery Engine will use tech from startup TAIGER

With an immense responsibility in accurately and efficiently matching solvers to problem owners, the OIP will be revamped with a new search functionality, dubbed the ‘Discovery Engine’. The feature is built using Omnitive Search, a solution designed and developed by Singapore-based startup TAIGER. The solution leverages artificial intelligence (AI) technologies including Natural Language Processing (NLP) and Ontology. These techniques also form the core of TAIGER’s AI offering in its automation tech stack, Omnitive.

What are Natural Language Processing and Ontology?

More specifically, NLP is a branch of AI that allows machines to read and make sense of human language to then be able to perform human tasks. Ontologies (aka Knowledge Graphs) on the other hand are machine-readable representations of subject-specific entities, such as those within the domain of solution provider capabilities. The ontology describes entities and their relationships with one another very expressively, almost as if mimicking the abstraction of ideas in a human mind.

“NLP and Ontology from the technological powerhouse of Omnitive. Together, these technologies work synergistically to form a robust search engine that perfectly understands natural language used in complex search queries and results, thereby matching appropriate pairs more effectively” said Guillermo Infante, Chief Technological Officer at TAIGER.

An Ontology graph of solution provider, TAIGER, and its related entities

Inside the Discovery Engine: how it works

So how exactly does the Discovery Engine work? After Problem Owners and Solvers upload their Challenges or Solutions, TAIGER’s NLP and Ontology technologies further read and index listings to refine their scope. Text contents from solution proposals are ingested and automatically tagged based on Omnitive’s Ontology, giving them machine-readable meaning and context. At the same time, text contents from problem statements are passed on as search queries that are processed using NLP.

The Discovery Engine then matches problem to solution and generates a list of search results based on their relevance. Results are further ranked based on various factors, such as whether the solutions have been used for similar problems previously. The engine is also smart enough to guide Problem Owners through a fallback process flow if it detects few or low confidence matches; this gives applicants guidance in refining their problem statements with minimal human intervention.

“Making innovation happen requires a precise match between the right solution provider and the right problem owner. But with the data within unstructured proposals and problem statements so difficult to parse, conventional search engines would struggle to find accurate search results. This is why the new features in the Discovery Engine plug a very critical gap in innovation discovery by eradicating issues with findability or discoverability between problem owners and solvers”, said TAIGER’s Head of Government Sales, Jaron Ong.

Process flow for solution providers and problem owners on the Discovery Engine

Better content discovery to unlock innovation at speed and scale

Minister for Communications and Information S. Iswaran shared in an update that over 10,000 companies are open to offering tech solutions on the platform—a number that will soon be easily managed with Omnitive’s Discovery Engine. Its installment will prove valuable in saving time and effort required to manually evaluate solution proposals. This is especially advantageous considering the complexities of the niche domain of crowdsourced tech discovery.

In a Straits Times feature, IMDA said this structured process “will support companies requiring more complex solutions to meet their innovation and digital transformation needs”.

With the new features implemented well, IMDA’s platform has high promise in expediting innovation at speed and scale, as a catalyst of new business opportunities both locally and abroad.

Read the full feature on the Straits Times