Search and Recommendation
Data is growing on an exponential basis with the largest growth in the form of unstructured and semi-structured data. The challenge for many firms is how to extract the value from these sources without relying on manual efforts. The EKG is a place where data and meaning come together, as “things-not-strings”, creating a contextually aware data environment to allow smarter search and recommendation services.
• Smart Search
• Chatbot
• News
Problem
Transforming the information embedded in unstructured text into meaningful and machine-executable content represents an untapped source of value. But the processes around identifying, managing and linking unstructured data with structured data systems are nascent, and the volumes are rising exponentially. Important information about customers, products or markets is buried and making these connections is vital to uncover insights to keep your business ahead of your competitors.
Impact
Not having all your data as machine-readable and executable will slow down your digital transformation efforts. Being too reliant on manpower that cannot scale as fast as the data is growing. You may also miss key insights that your competitors find if you don’t have access to the best, most highly connected data.
Solution
The EKG acts as the mechanism to connect structured data with unstructured content, and the EKG ontologies provide contextual references to support smarter searches and also aid the natural language processing (NLP) technologies which translate words to data. Substituting manual human-oriented processes with AI-assisted functions enables more accurate results by combining meaning with machine-executable business rules and graph analytics.
Smart Search
The research department of a global bank has many thousands of unstructured documents making it hard and time-consuming for analysts to locate relevant content for a class of financial products, a single company or industry sector. Documents are scanned using natural language processing (NLP) techniques and taxonomies from the EKG to extract important business concepts – resulting in improved classification and more relevant answers for their analysts.
Chatbot
A retail bank creates a smarter automated customer service assistant to help business users answer questions quickly and more accurately. The EKG is used by the chatbot service which parses natural language, uses ontologies to match similar terms and best answers from other customers - to recommend the most likely response. The customers get better quality and faster response time while the provider reduces load on their call center.
News
A financial data provider uses structured and unstructured data from many sources to gather news and announcements about companies and their securities. Using NLP techniques, a domain ontology and data from the EKG – they can efficiently process corporate actions, announcements and automatically cross check quality against other sources. The improvement in both accuracy and efficiency allows the team to focus on exceptions and frees SME resources to concentrate on more value-added tasks.