Quintessentially search, no matter what the terminology

Semantics: the search for meaning

Whether information is received as a result of searching by the user or proactive delivery through an application, the goal is to understand its meaning and relevance. Semaphore is a semantic platform from Smartlogic that auto-classifies information assets using taxonomies and ontologies and then enriches them with metadata. That process enables a search engine to interpret the user’s intent more accurately and therefore produce better results.

Building a semantic model is an important component of Semaphore. Semantic models can be used to disambiguate diverse vocabularies, which is critical for self-service. “Companies are moving aggressively toward self-service,” says Anne Lapkin, SVP for global strategy at Semaphore, “but in order for that to work, the system must translate the user’s input into terms that allow for retrieval of the right information.” So if a customer states that the “blue screen of death has appeared,” the equivalence to “system crash” must be established. With an understanding of the relationship between the user’s input and the content in the knowledgebase, a search engine can present the right solution.

“Another example of the value of semantic models is their ability to make sense of data that comes from a sensor and is stored in a smart data lake,” Lapkin explains. “The numbers as they are sent from the sensor to the database are not meaningful. But through the model, they can interpret them to provide output that is meaningful to users and discover, for example, that the sensor readings are out of range for a device and that it is likely to fail.”

Semantic platforms provide the mechanism through which inferences can be drawn even when specific words are not used. That capability is particularly useful in ensuring compliance—for example, in protecting personal information or enforcing anti-money laundering policies. “Some types of information are now considered sensitive that were not before,” Lapkin adds. “For example, in some contexts such as the EU’s new General Data Protection Regulations (GDPR), political affiliation needs to be protected.” Therefore, information such as attendance at a fundraiser might need to be kept private, even if a political party was not identified.

An underlying model in Semaphore allows harmonization of different vocabularies that are being used for such applications as clinical trials. “If two organizations are conducting trials for a pharmaceutical company, they may use different formats and terminologies,” Lapkin explains, “but at some point, the information needs to be integrated. By establishing the equivalence among different concepts and terms, Semaphore allows information to be searched and summarized more easily.”