By Subrata Chakrabarti, VP Marketing, Kyndi
Employee productivity is the lynchpin to business agility. But employees still struggle to efficiently find the key information they need. That’s because 80% of enterprise data is unstructured. Much of it is in free-form text in reports, research, emails, and other documents—scattered around the enterprise. Looking for parts of information hidden inside a manual, discovering exemption clauses in a legal contract, or locating specific policy information at a country level – requires access to answers from such unstructured and untagged content.
A recent IDC report suggests that the average knowledge worker wastes 10 hours or more per week looking for information from documents and manuals. In the past, organizations opted for keyword-based search techniques. However, with growing volume and variety of unstructured content, keyword or term-specific search techniques began to struggle. If the users don’t know about a specific search term, such as a machine part name or type of legal contract, the search system fails to deliver since it doesn’t know where to look for it. The result is growing employee frustration and lost productivity even when the enterprises are spending more money in data initiatives.
Enter cognitive search – an approach that goes beyond the literal interpretation of words. When we use natural language, such as English, we have the choice of expressing ourselves in different ways – using words that may be semantically similar, synonyms, short and long forms, and even acronyms. We may also want to drop in industry jargons or terms that are heavily used within our specific organization. Wouldn’t it be easier if a search technique could automatically understand them – as if, it could read our mind? What are some of the capabilities that we would want specifically from such an experience?
Understand the user intent behind the search terms: Go beyond the literal sense of the words and comprehend the grammatical construct, expressions, entities, and key concepts.
Allow for resiliency and variation in user queries: The search experience needs to be powered by automatic support of language variations – for example, semantically equivalent words, similar word forms and lemma variations within the language, acronym resolution, and synonym support.
Be forgiving in user engagement: The search technique should automatically resolve typos and word misspellings, as well as account for short and long form variations of different words.
Customize easily to fit domain and company needs: There should be robust support of custom terminology and domain-specific terms.
Empowering continuous improvement: Clear explanations of the results need to be displayed to help improve search performance. Such a capability could enable enterprise developers to fine-tune for enhanced performance.
The human thought process is not linear. Finding the right answer requires thoughtfulness, creativity, discovery, and iteration. Consequently, business users need a search solution that helps them think and discover more efficiently. This is what a next-generation cognitive search is all about.
What value would it bring to your enterprise today if you could use a cognitive search solution that you can implement quickly? How much time would it save by automatically analyzing long-form text, manuals, and other documents to find actionable insights for you quickly? And also do this in a smarter, faster, and more explainable way than anything you have used before. How much more engaged would the employee experience be as a result?
Contact Kyndi to discuss how to get started on a game-changing cognitive search solution, powered by AI, which works by understanding the user intent.