Semantic technology:
From sentiment to applications

Hakia's innovation moves beyond sentiment analysis, which is useful in customer support and advertising. Hakia's approach, according to the Semantic Technology Blog, pushes ahead by "building a recommendation model on top of semantic filtering of the content ... Its delivery model is subscription/consumer-based, which is different from most of the ones that have tried this before or are trying now, as they were/are more enterprise-focused with the exception of StockTwits."

One should weigh some considerations before getting rid of a personal financial adviser. For example, the service requires a subscription. Details about the inner workings of the SENSEnews system are sparse. The latency between the availability of information and the updating of the SENSEnews index is not evident.

Three changes

As I look over the uptake of semantic technology in the enterprise and in consumer markets, three changes are taking place. First, there is a growing awareness that key word search is not appropriate for certain types of information retrieval tasks. A few years ago, the notion that a search box would unlock the treasure chest of high-value information was exhausted. Users grew tired, then annoyed and understandably more vocal in their criticism of search systems that generated a need for endless opening, scanning and processing of "hits." When most of the items in the search results list were irrelevant to the user's information need, a semantic window opened. Attensity and Hakia are two firms trying to squeeze through that aperture.

Second, the semantic technology is not exposed. The systems and methods are manifested in an application that solves a problem or does something the user can look at, evaluate and decide if the output has value. Equations, canned demonstrations and lectures about latent semantic indexing lose sales in my experience. What is encouraging about Attensity's push into Madison Avenue and Hakia's foray into stock picking is that the outputs are useful and concrete. Today, unless "there's an app for that," semantic technology will be just another content processing complexity. Made fungible, semantic applications can be magnets that pull customers.

Third, the semantic application market is just now taking shape. Are Attensity and Hakia going to remain static? No. Both firms will continue to evolve. In addition, the attention both attract will create the now familiar cycle of what I call the sheep approach. Innovators with semantic technology will follow Attensity, Hakia and others into the market for semantic applications. This is exciting on one hand and a repeat of the all-too-familiar pattern of competitor proliferation. For those who are stakeholders in Attensity and Hakia or other companies working in this space, a buyout may be in the future.

I am not ready to say "a semantic revolution" has occurred. But it looks to me as if there are signals that traditional enterprise search, content processing and knowledge management solutions will face a choice: Go semantic or go away.