Turn Customer Service Calls into Enterprise Knowledge Graphs
The need for text analytics and speech recognition has broadened over the years, becoming more prevalent and essential in the sales, marketing, and customer service departments of various types of businesses and industries. The goal is simple for these contact center use cases: provide real-time assistance to human agents interacting with potential customers to close sales, initiate them, and increase customer satisfaction.
Until fairly recently, the rich array of unstructured data encompassing client texts, chats, and phone calls was obscured from contact centers and organizations due to the sheer arduousness of speech recognition and text analytics. When readily integrated into knowledge graphs, however, these same sources become some of the most credible for improving agent interactions and achieving business objectives.
Powered by the shrewd usage of organizational taxonomies, machine learning, natural language processing (NLP), and semantic search, knowledge graphs make speech recognition and text analytics immediately accessible, enabling real-time customer interactions that can maximize business objectives—and revenues.
Taxonomies are the foundation of the knowledge graph approach to rapidly conveying results of speech recognition and text analytics for timely customer interactions. Agents need three types of information to optimize customer interactions: their personas (such as an executive or a purchase department representative, for example), their reasons for contacting them, and their industries. Taxonomies are instrumental to performing these functions because they provide a hierarchy of relevant terms to organizations.
In sales, simply by ingesting respective taxonomies for sales-related terms and products into knowledge graphs, organizations have the blueprint to quickly analyze documents, phone calls, and chats via text analytics. These taxonomies are useful for extracting entities from this data, forming the basis of sentiment analysis, and facilitating several key metrics for increasing agent performance. Organizations can compare how much sales representatives discuss certain products or categories of products, how much time they spend with customers, or any other metric for dynamic routing and other value-adding applications.
Natural Language Entity Extractions
Knowledge graphs’ taxonomy-driven speech recognition platforms utilize NLP for entity extractors that parse text according to taxonomies, uncovering pertinent business concepts. Although speech recognizers transfer voice content to text via NLP, taxonomies are still required to account for nomenclature such as product names. Once this text is input into graphs, NLP-powered entity extractors use taxonomies to collect the relevant terms for business metrics.
Positive and negative words are isolated for sentiment analysis, and specific terms are extracted to determine the nature of interactions, the technology referenced, and the potential client’s role and industry. Since this information is stored within knowledge graphs alongside other customer and product data—including external data sources like social media feeds—organizations can perform startlingly specific queries about customers and agents. These analytics let them learn which products are most popular for whom and why, and best practices for tailoring agent interactions to exploit this knowledge.
Machine learning helps organizations and their clients create optimum interactions with speech recognition analytics in two ways. The first is during training for text analytics models, which requires labeled input data for supervised learning. Once text, entities, and taxonomies are stored in knowledge graphs, people must label individual chats or calls according to the aforementioned categories. The subsequent labels serve as annotated training data for classification models so when new calls are received, they’re accurately classified according to labels to inform agents of the factors that are most meaningful to individual callers.
The second way machine learning empowers speech recognition systems is by identifying patterns useful for satisfying customers. This powerful technology employs statistical analysis to see which topics correlate the most, which agent had the most success with which type of approach for which type of client, and other data-driven patterns useful for recommendations. Thus, agents will know that if customers reference a specific type of router, for example, the former should mention a certain security system, too, because machine learning indicates these products correlate with the customer’s persona.
In addition to customized recommendations, business metrics for agents, and strategies for dynamic routing, taxonomy-driven speech recognition knowledge graphs also provide semantic search. The combination of sentiments, classifier models, and entities enables organizations to semantically explore their unstructured voice and textual data with search capabilities that understand users' underlying intentions. This final advantage is ideal for both tactical and strategic use, particularly because semantic search in knowledge graphs can also encompass an array of additional knowledge internal or external to the enterprise. It’s the final step to mastering data mining of unstructured text analytics, and it can prove invaluable to any organization struggling to leverage this data for competitive advantage.
Jans Aasman holds a Ph.D. in psychology and is an expert in cognitive science as well as the CEO of Franz.com, an early innovator in artificial intelligence and provider of Semantic Graph Databases and Analytics. As both a scientist and CEO, Aasman continues to break ground in the areas of artificial intelligence and semantic databases as he works hand in hand with organizations such as Montefiore Medical Center, Blue Cross/Blue Shield, Siemens, Merck, Pfizer, Wells Fargo, and BAE Systems, as well as the U.S. and foreign governments.