Semaphore & MarkLogic: Driving benefit search in healthcare

Download Case Study (PDF File)

Full Story


Driving consumer satisfaction comes down to delivering the right information at the right time. As healthcare becomes consumer-centric, healthcare organizations need to create systems that empower consumers to easily engage with all aspects of their care. For healthcare payers, this means developing a holistic view across a diverse set of information sources, such as benefit plans, provider services, patient information and claim data. As healthcare organizations transform from retrospective, reactive service enterprises into proactive engagement platforms, they need a flexible data platform that provides a comprehensive view across these key information sources.

A top 10 healthcare payer needed to provide their customer support team with an efficient and reliable mechanism to respond to member queries about benefits and coverage. While each member is associated with a single healthcare plan, plans are often written by medical and legal experts who don’t use the same vocabulary as a user. This diversity of language makes it difficult for the customer support team to quickly identify and respond to member queries. This payer built a Smart Search strategy that combined MarkLogic’s database and Semaphore Semantic AI platform, to improve customer support, reduce operational costs and position them to provide timely and innovative customer self-service solutions to their customers.


MarkLogic is a fully transactional and operational database platform. Rather than worrying about time-and resource intensive ETL (extract transform load) processes characteristic of traditional relational technologies, MarkLogic uses a flexible, multi-model approach that can easily handle almost any data source, whether it is data from existing relational databases, mainframes, Hadoop, file systems, or any other source. As an enterprise ready platform, MarkLogic also offers transactional reliability, high availability and disaster recovery, and certified granular security at the document level.

Semaphore Semantic AI platform is an enterprise grade semantic solution that integrates with content management systems, workflow and search engines and next generation databases like MarkLogic to solve complex problems that traditional technologies cannot. Semaphore brings structure to the unstructured, scales to manage organizational volumes, and supports industry standard semantic vocabularies.

Semaphore leverages knowledge models and sophisticated semantic strategies such as AI, machine learning and natural language processing, to enrich assets with precise, complete and consistent metadata. This allows organizations to aggregate and link disparate data sources, extract value from information, make sense of data and identify patterns and relationships to drive insight. Metadata is a key component in the Smart Search strategy, it provides the context and meaning to transform simple key word search results into a robust search experience.


They began by using Semaphore Text Miner’s extraction algorithms to identify common concepts found in the benefit documents. The mined concepts were loaded into the model, enriched and stored as a graph in the MarkLogic database. Semaphore Classification and Language Services uses the model and applies sophisticated natural language processing, entity, fact and relationship extraction techniques to tag the benefit plan documents with precise, complete and consistent metadata, which is stored with the benefit plan documents in the MarkLogic database.

When a user performs a search, the search term is first located in the model graph, and then expanded by incorporating the concepts that are semantically near (i.e. synonyms, acronyms, alternative spellings broader and narrower concepts) the search term. MarkLogic’s search process maps the expanded keywords to the appropriate benefit plan document sections and returns the relevant results. Combining Semaphore’s benefit plan classification and MarkLogic’s keyword expansion, provide customer support with “Smart” search results.

Today, customer support staff can quickly and accurately respond to member requests; a search for “morning sickness” can easily locate coverage information from the maternity benefits section of the policy. With Smart Search, support costs are reduced, staff and member satisfaction is increased and the organization is positioned to extend the reach of their Smart Search platform to additional use cases within the enterprise and support customer self-service delivery.

Learn more
about the products


Keep exploring
stories like this one

Read Next Story