Long gone are the days when a simple search box was all that was needed for an intranet to be effective.
With the wealth of data and information that employees need to retrieve to do their jobs effectively, organizations must have the most innovative tools and technologies in place to offer not just any search results, but search results that deliver the most relevant data — data that can support better insights and decision-making. In the past few years, big data analytics, machine learning and artificial intelligence (AI) have evolved rapidly, allowing enterprises to better use their intranet data to produce search results that are more insightful and valuable.
We’re already seeing the impact of AI in the digital workplace and, in its November 2017 report “Predicts 2018: Artificial Intelligence,” (fee charged) research firm Gartner argues that “by 2022, 40 percent of customer-facing employees and government workers will consult daily an AI virtual support agent for decision or process support.”
Weak AI for Strong Enterprise Knowledge and Collaboration
How can AI improve intranet search effectiveness? And how does it work? It starts with semantic search along with natural language processing (NLP), machine learning and big data analytics. These technologies are helping enterprises build AI-assisted tools like question-answering systems and chatbots that can play integral roles in the new digital workplace.
First, let’s clarify what type of AI we are talking about. Experts often break down AI deployments into “weak” AI, which are technologies that simulate human behavior, and “strong” AI, which are technologies with the qualities of “consciousness” and the capability for original thoughts.
In the case of improving business operations with real-time analytics dashboards, chatbots or question-answering systems, we are talking about a form of weak (or limited) AI. But even though these tools are “weak,” they can still make for a much stronger corporate intranet and deliver immense value by automating and improving functions like document search, information retrievals, customer support and daily administrative tasks.
Related Article: AI-Driven Enterprise Search Is Closer Than You Think
Semantic Search, NLP and Knowledge Graph
More and more people want to leverage AI for better corporate-wide search, which leads them to semantic search, NLP and the knowledge graph.
NLP started out in the 1950s in the world of computer scientists. But now, thanks to tools like Siri, Alexa, Cortana, Wolfram Alpha, Watson and Google Home, NLP-powered applications are starting to make their way to the enterprise, delivering new ways of interacting with data and finding information.
We’re seeing NLP applications being developed with the intention of tailoring language and knowledge models to the unique domains that are found in the enterprise. This type of end-to-end question-answering system that lives on top of search is truly a semantic extension to the search box with far-reaching implications for all types of search, including corporate intranet, ecommerce, website, publishing and customer support.
Machine-Readable Knowledge Required
These NLP systems are machines that understand human language and human endeavors. They can converse, to a limited degree, with humans about topics that both parties understand. But getting there is going to be half the battle.
For a machine to be able to converse with a human, it must have machine-readable knowledge. Thus, defining practical approaches to encoding human-readable to machine-readable knowledge will bring tremendous advantages to intranet applications. For example, a chatbot powered by NLP can answer simple but high-volume support or fact-based questions, reducing the amount of time employees spend answering customer questions and freeing them up to focus on more strategic tasks.
But NLP has other enterprise applications beyond customer support. Imagine if, rather than having to sift through multiple reports, spreadsheets and documents and do your own compilation, you could get immediate answers simply by typing in or asking questions like these:
- How many new customers did we acquire this month?
- How does the North America’s revenue compare to Europe’s revenue this year?
- How do I request office supplies?
This is happening today with the emergence of the knowledge graph. The concept of the Knowledge Graph began with the launch of Google's Knowledge Graph in 2012. The goal was to provide answers to queries beyond a link by accumulating information from a variety of sources adding to its knowledge over time. Knowledge graph is now used to refer to any system doing similar work.
To implement this type of system, you create knowledge bases that are used to enhance search engine results with information gathered from a variety of sources. The knowledge graph is powerful because it contains the knowledge base specific to your business domain. Therefore, when coupled with NLP for semantic search and question-answering, it delivers information that is highly relevant and immediately useful.
With some of these practices put into place, we will see AI technologies bridging the gap between the user and intranet data, improving the employee or internal user experience. If planned and implemented correctly, tools like question-answering systems, chatbots, cognitive or intelligent search systems, and AI will help improve the user experience within the enterprise, solve information retrieval problems, lower costs and increase revenue.
Learn how you can join our contributor community.