A recent report from Gilbane Group senior analyst Lynda Moulton makes good reading for enterprise business leaders who want to better understand why and how their organization can â€“ and should â€“ consider semantic software technologies as core to turning the vast amount of information business units possess into knowledge that they can act on.
The paper, Semantic Software Technologies: Landscape of High Value Applications for the Enterprise, has as its focus helping enterprises identify, select, and implement the appropriate software for work functions (marketing, customer services, BI, content management, enterprise search, e-discovery for R&D, etc.).
There’s a ton of electronic data about that can surface major business impacts, in areas ranging from discovery of business developments such as competitive shifts or learning about market trends that present opportunities for growth, but “it has become impossible for humans to adequately survey the vast electronic information domains accessible to them in a reliable and timely manner,” Moulton writes. “The quantity of structured and unstructured content, plus its lack of meaningful linguistic uniformity, has overwhelmed any human process previously used to extract value.”
Foresighted business planning calls for consistency in vocabulary and content practices, she notes. And to that end investing in ensuring the right answers, the best information, and efficiency of processing is paramount for industries “when content is sizable (millions of documents), complex in scope and depth, high-value to narrowly focused audiences seeking only small portions out of the entire corpus, or needed by experts for use in their areas of expertise,” as well as where there is “obscure content with importance to niche audiences can be brought to light for a creative research.” Life sciences and publishing lead, but others will follow.
The paper provides definitions and background of technologies, from text mining and analytics to NLP to sentiment analysis, to prepare readers to understand what to them is still a new type of business software, and follows that with go-forward adoption guidance â€“ who are the players and how specifically are their applications used, and how to select and implement a product. This is useful reading not just for the business professionals involved, but of course for technologists who can play a role in proof of concept efforts and those IT leaders who will be active in actual deployments.
The paper is sponsored by vendors in the semantic technology space, including Cognition, Expert System and Linguamatics, which get a Deep Dive profile outside the main research report that covers the overall market in a balanced analytical fashion. Indeed, the author clearly points out that, “The task of defining semantic software technologies is risky because this is a new field in which no one technology method or model has reached a position of clear market leadership,” and that “There are no companies or products in this field that have eclipsed all others, as offering universal semantic processing or semantic search, yet.”
It’s All About the Team, and the Teamwork
There’s still time to be an early adopter â€“ and that means there’s also time to ensure that whatever semantic software option an enterprise concludes is worth its investment is intelligently implemented and supported, according to the research.
So, one of the interesting things about this report is that it delves pretty extensively into how to structure teams to select and apply semantic software, drawn from Gilbane’s interviews with subjects who’ve managed or championed semantic technology initiatives. This is information that’s gold for enterprises getting their foot in this door â€“ especially in its cautions that it’s wrong to assume that team expertise can begin and end with addressing the technology foundations and scope. Using the applications â€“ how humans interact with semantic technologies â€“ has to be part of the team’s consideration, as well. “Extracting concepts when language is complex or foggy is a huge semantic challenge that technology alone cannot solve easily,” Moulton writes. “There are still plenty of areas in which humans are needed to untangle complicated language and this is one of them. It serves as an example for those who believe that installing a software application is the solution to anything.
Gilbane’s research shows that team core competencies that matter are backgrounds in computational linguistics, informatics, subject matter experts in the domains being targeted, taxonomists, and use experts with enthusiasm and motivation to get better results from search. But Moulton notes that the newness of informatics and the use of semantic software means it’s okay that most team members will bring little direct experience with these technologies to their work. As long as they have some familiarity with the subject discipline, multi-disciplinary backgrounds, and are conceptual thinkers with an orientation to systemic modeling and problem solving, an enterprise should be off to a good start. That is, as long as they also “sort out the issues of ownership and leadership [between IT and proof of concept teams] before making any moves toward acquiring semantic software technologies. It is complex enough without getting into turf battles over its selection and implementation.”
If semantic software is on your enterprise’s radar â€“ or if you are inclined to think that it should be â€“ take a look at this research. It’s a good read and stands to be a big help.