I am pleased to see Paragon included in the Semantic Knowledge Graphing section of the Gartner Hype Cycle for Life Sciences 2013 this past summer. Being technology agnostic, we believe semantic knowledge plays a key role in how innovation can be truly explored and harnessed. As connectivity for all stakeholders is increased, it also underscores the need for business leaders to see how new technology approaches can be used as Innovation management becomes part of their daily responsibilities.

Semantic Knowledge Graphing

Analysis By: Michael Shanler

Definition: Semantic knowledge graphing is software with supporting services that enable researchers to search, mine (chemical structures, text, system biology models and so on), aggregate and share complex life science data relationships. This includes connecting the text from journals, chemical structures, biomolecular content, disease pathways, and other clinical, research, and scientific relationships. These systems map internal and external datasets, have the capability to continually search for updated information, and support synchronous sharing.

Position and Adoption Speed Justification: This is a new profile in 2013. Previously, parts of it were covered in other Hype Cycles over multiple profiles, such as text mining, semantic search, analytics, visualization, big data, electronic laboratory notebooks and scientific informatics. Knowledge graphing has arrived in many different forms; however, it was only over the past few years that the applications and infrastructures were injected with the semantic search capabilities and graphical relationship models necessary for handling scientific big data, with the massive computing power to improve performance.

Knowledge graphing is typically performed in smaller labs by specialty data scientists. Until recently, mapping software was designed to tap into only a few of the many available data sources (such as internal data warehouses, internal data marts, application silos, subscription databases or data from the public domain). The precursor systems were never designed for performance when handling large datasets, and they suffered from severe performance limitations due to a lack of computing power and poor orchestration.

There is a steep learning curve for using this software, which is slowing the adoption rate, and few users within organizations will know the R&D and the IT requirements. As companies feed the desire for expanding and sharing scientific knowledge sets, these systems will become easier to use and have higher performance, so clients will expand their adoption. In the next two to three years, the infrastructures for handling big scientific datasets will evolve and make the ROI on knowledge mapping systems easier to realize.

User Advice: R&D IT should explore these systems with the goal of improving knowledge mapping and collaboration by developing insights from complicated scientific big data. The conversation on ROI will involve strategic R&D heads in addition to IT. Since there is a high level of complexity associated with learning these systems, IT must partner with scientific leads, data scientists and informaticians to develop internal best practices for their use. IT should also use the same individuals to address data quality, data standards and common ontologies. In addition, before building out new systems to connect internal and external data that is relevant to R&D, IT should work with department subject matter experts to understand which datasets need to be connected. Finally, it is important for end users to be educated on the different aspects of big data, because the volume, velocity and complexity will dictate which systems deserve investments.

Business Impact: The use of these systems can help accelerate innovation activities, disseminate complex relationships with scientific stakeholders, and support collaboration and innovation strategies as they relate to drug discovery, translational medicine, competitive intelligence and clinical research. Interpreting the data and analyzing the results of these systems are complicated processes, and will initially tax R&D IT resources as IT bridges the software, service and project spaces.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Adolescent

Sample Vendors: Cambridge Semantics; Cytoscape; Dassault Systemes; Entagen; Paragon Solutions; Thomson Reuters

View the entire list of Hype Cycles on Gartner's Blog.