Curt Monash Analyzes the Text Analytics Market

Curt Monash recently shared his views of the text analytics market through his blog and a slide presentation that he’s made available online. The presentation is refreshingly hype-free, and I recommend you take a look.

His observations about the web search market are spot-on: the current attention is on transactional queries (see Andrei Broder’s classic paper on the taxonomy of web search for an explanation of navigational, informational, and tra…

Curt Monash recently shared his views of the text analytics market through his blog and a slide presentation that he’s made available online. The presentation is refreshingly hype-free, and I recommend you take a look.

His observations about the web search market are spot-on: the current attention is on transactional queries (see Andrei Broder’s classic paper on the taxonomy of web search for an explanation of navigational, informational, and transactional queries), and web search generally is dominated by the dynamics of adversarial information retrieval. Depressing (to me, at least), but accurate. He does see potential future with better interfaces, but he asks, “how good does the technology have to get before people care?” My sentiments exactly.

On to the enterprise market, which is more interesting. Here it’s harder to summarize Monash’s thoughts, except to say that he sees the current landscape of enterprise search offerings as hopelessly confused.

Monash divides the enterprise market into public-facing site search, which he further divides between e-commerce and “general”; and “true” enterprise search which seems to mostly denote intranet search; and custom publishing. While I’m not entirely comfortable with his taxonomy of the space, I do give him credit for laying one out.

He then goes on to explain how “one-size-fits-all” approaches have failed and how the enterprise search market landscape is “bollixed”. He lists a number of technical challenges, all of which I agree with.

But I’d add one: the need for content enrichment techniques and interfaces that support interaction, exploration, and discovery. Yes, we’ve seen these as buzzwords in vendor hype, but that doesn’t make them any less real. There’s been too much emphasis on best-first, known-item search, and not enough on the other use cases that comprise enterprise search and information access.

I think that exploratory search will eventually be important for web search too, but the complacency with current approaches kills any sense of urgency. There is no imminent threat to Google’s reign.

In the enterprise search market, however, there is a justified dissatisfaction with the status quo. And, in my belief and experience, that is because too many people (vendors and enterprises) are trying to treat the enterprise like a microcosm of the web, where the only major differences are the connectors to acquire content and the ranking algorithm to sort results. Getting these right is necessary but not sufficient. Interaction, exploration, and discovery–in short, HCIR–are not just nice-to-have features, but rather are essential to making search work in the enterprise.

Link to original post