We haven’t yet worked out all the issues that will enable the Semantic Web to gain traction on a mass scale across applications, processes and industries. But enough of them have been addressed so that we can begin to understand what portion of the Semantic Web can support real commercial opportunities. That’s how Mark Greaves, director of knowledge systems at Paul Allen’s asset management firm Vulcan Inc. sees it – and that has positive implications in particular for three market areas.

These areas include enterprise BI, and more broadly strategic enterprise IT; web marketing; and newer web 3.0 businesses, says Greaves, who will be speaking on the Semantic Web’s evolution at next week’s Web 3.0 conference in Santa Clara, Calif. Greaves has been a key part of that evolution, both with the work he currently has underway at Vulcan’s Project Halo, and in his past life as program manager in DARPA’s Information Exploitation Office for the DAML program. DAML developed many of the initial languages and tools that lie at the foundation of the Semantic Web. Here’s Greaves’ take on both the opportunities – and the challenges – that surround the Semantic Web today.

● Market #1: Enterprise BI
In the enterprise, Semantic Web and Linked Data technology can enable BI systems to resolve a much wider range of queries, Greaves says. There are many seemingly straightforward queries that are impossible to execute without human intervention, often simply because necessary data isn’t maintained in a place that is accessible to the BI system. “Semantic technology allows these queries to be resolved semi-automatically by accessing knowledge available on the web,” Greaves says.

For example, suppose you want to know the number of widgets you’ve sold to cities with populations of more than 500,000 headed up by Democratic mayors. “With semantic technology, including some of the new linked data techniques that are coming on line, you can answer that kind of query by dynamically linking to web-based data repositories to retrieve city population data and mayoral party information,” he says – without needing to have a pre-existing data relationship with the owners of those repositories.

Semantic Web standards such as OWL, RDFa, and Linked Data reduce the requirements around contacting owners of that data, and around implementing the schema translations that enable the integration of distinct sets of information across organizational boundaries. “One of the things this technology makes possible is doing web-based queries without pre-existing data sharing relationships, in the same way that we don’t need a pre-existing relationship if I decide to hyperlink to your web page from my web page,” Greaves says.

While companies are understandably reluctant to publish their proprietary data on the web, there are important cases where businesses can gain cost advantages by joining together to share critical but non-proprietary data. An example Greaves cites is the Linked Drug Discovery initiative, where several large pharmaceuticals companies are starting to leverage the Semantic Web techniques to link together foundational information that is needed by all of them for drug research, but which is not particularly proprietary to any of them.

● Market #2: Marketing and Semantic SEO
Semantic technology, including Semantic Web technology, also provides a real opportunity in marketing – “in particular a very dynamic area of marketing, which is search and presentation on the web,” Greaves says. He notes that Yahoo recently reported up to 15 percent higher click-through rates for richly formatted snippets, “and the driver for these rich formatting systems is various forms of semantic coding embedded in page HTML. There are several different kinds of semantic coding techniques – RDFa and microformats are the main ones—but the key commonality is the way that these techniques surface the semantics in the crawled HTML pages in order to drive rich snippet presentation in the search engines.”

● Market #3: Web 3.0 Businesses
Greaves believes Semantic Web technology can support a new set of web 3.0 businesses based around curating and exploiting large distributed web-based knowledge bases. Metaweb (creators of Freebase) and Vulcan-backed EVRI are examples of companies that use different ways to synthesize high quality web-accessible knowledge bases out of lot of different data, harmonizing and semantically indexing it, applying quality control, and then using it to enable a more data-rich framework in which to support the kinds of activities that people want to do on the web. These include creating content, seeking information, being updated on current events, socializing, and making purchases. “Web sites that support those activities are hungry for rich data to really drive context and personalization for their users, and semantic technology—particularly the kind EVRI and Freebase and their competitors specialize in—can be used to efficiently support that data infrastructure role,” Greaves says.

Challenges Around Opportunities
These three markets share many of the same challenges, Greaves says. “There are major open questions around ensuring data quality on the web -- not just spam-type data quality issues but also about the fundamental quality of markup and the quality of semantic inferences that can be made off of that markup,” he says. Those issues are beginning to be addressed by the wider use of ontologies that are derived from standards bodies, and by the fact that “the same kind of trust networks that underlie the current web are starting to emerge in the Semantic Web,” he believes. “Data quality will be a long-term issue that is never 100 percent solved, but it will be solved to a level that will enable commercial usage.”

The other issue, he says, has to do with expressivity -- whether the formal languages and technologies that underlie Semantic Web can adequately express the information that people wish to receive. Take, as an e-commerce example, using Semantic Web technology to express details about transactions concerning a product. Price information is straightforward to encode, but it is more difficult to express all the rules and policies that might obtain around delivery, shipping insurance, warranties, contracts, returns, and how each of those pieces might differ across states or countries, for new customers or long-standing clients, and so on. Ontologies such as GoodRelations are excellent but they generally don’t delve into this level of detail. “GoodRelations may indicate there is a policy on product returns but not what it is,” Greaves says, but many in the industry (including Vulcan) are working on addressing this issue.

In this economy, one also is prompted to ask whether the slowdown in VC funding is another challenge to the realization of the opportunities that exist around the Semantic Web. Says Greaves, “I think the VC community is appropriately skeptical about this technology, as it should be about every new technology. But I think that by forcing entrepreneurs to really focus on the business values of the semantic technology, VCs are performing a valuable service.”