Mine Enterprise Information through Cognitive Virtual Assistants
Those who seek may not always find what they are looking for, more so in the case of enterprise information. According to Forrester, 54% of global information workers are stonewalled multiple times a month when they search for information, insights and answers. This negative impact on productivity is profound. Making information more accessible is not as easy as it sounds.
What hampers information access?
Enterprise content–whether that is website content, marketing collateral, user product guides or policy documentation–is typically the most underused and overlooked asset of an organization. Part of the reason has to do with how enterprise content is managed. More often than not, enterprise content is dispersed throughout organizational silos, making it difficult to access information. This “siloization” happens for three reasons:
1. Cultural: Aggressive competition between departments may prevent knowledge sharing and cross-pollination
2. Structural: In large organizations, each team owns and updates information that is specific to its department or team, but lacks organizational focus and cohesion
3. Technological: Because it lacks the technology and know-how, the enterprise is unable to connect these content nodes or silos
Content Management Systems (CMS) attempt to solve this problem, but are constrained by divergent writing styles and employees’ varying levels of commitment to keep the information accurate and timely. In addition, having various contributors is bound to lead to inconsistent information structures, unless all contributors are well-trained. One contributor may miss adding a call to action, and another might add information that is irrelevant–or worse inaccurate. In general, CMS makes it challenging to provide uniform advice.
The era of Cognitive Search
The emerging field of cognitive search allows enterprises to collect, organize and pull information from multiple sources – be it structured content in a database or unstructured content in documents, emails, images, or videos - eliminating the need to maintain consistent content. Powered by NLP and ML, Cognitive Search can understand the intent behind the query, gather essential content and respond with relevant results. This is a profound achievement in the field of Artificial Intelligence.
What role does the virtual assistant play?
The lines between Cognitive Search and Virtual Assistants are quickly blurring and diminishing. Today, cognitive interfaces are emerging and act as a single point of truth where knowledge resides. These state-of-the-art solutions can not only read and understand enterprise knowledge, but also unlock information to converse with and assist humans.
As opposed to stand-alone Cognitive Search solutions that only respond to a user, a Virtual Assistant can actually build conversations around the query and help people complete tasks in real-time. For example, imagine you’re a global organization that has 1,000 HR documents related to vacation policy–which varies by country. The Virtual Assistant is able to ingest that information and tell any employee how much vacation time they’re eligible for, how much time they’ve accrued and then provide the paperwork needed to complete an action. Any information can be mined and analyzed and any task can be executed through the virtual assistant.
Originally posted in Nasscom community page.