Ancient Search Recipes: Bread Pork Chops

December 16, 2019

I noted a report in the Times of Israel titled “Cache of Crypto-Jewish Recipes Dating to Inquisition Found in Miami Kitchen.” One of the recipes explained how to make a pork chop from bread and milk. (Dairy? Guess so.) Here’s what you and I can whip up using this ancient recipe:

image

The cookbook contains information which the author “didn’t think to question the idiosyncratic customs her mother and grandmothers practiced in the kitchen.”

By coincidence, my news alert spit out this article in the same list: “The Growth of Cognitive Search in the Enterprise, and Why It Matters.”

Magic. Bread pork chops created from zeros and ones.

Search matters. Cognitive search matters more. Who buys? The enterprise.

The write up recycles the equivalent of the break pork chop formula. Mix jargon, sprinkle with the notion of federated data, and bake until the checks clear the bank.

The article is fascinating, and it overlooks a few milestones in the history of enterprise search. What for example? Glad you asked:

  1. Forrester, the Wave folks, has created a report for its paying customers which reveal that search is now cognitive, able to tap dark data, and ready for prime time. Again! The Wave returns.
  2. Big companies are into search, including Microsoft  with its Fast inspired solution and Amazon Kendra with an open source how de doo to Elastic and LucidWorks. Some use old spices; others, open source flavoring with proprietary special seasonings.
  3. Outfits which have been around for more than a decade like Coveo are now smarter than ever in their decade long effort to pay off their patient investors
  4. Autonomy gets a nod despite the interesting trial underway in the UK.

The point is that enterprise search is going to be in the news whether anyone wants to revisit hyperbole which makes the chatter around artificial intelligence and quantum computing seem rational and credible.

Here’s a quick refresher about why untapped data in an organization is likely to remained untapped or at the very least not tapped by vendors of smart key word search systems:

First, data are in silos for a reason. No enterprise search system with which I am familiar can navigate the permissions and access controls required to put siloed data in one index. There’s a chance that the Amazon blockchain permissions system can deliver this, but for now, the patents are explanations and federated enterprise search is a sales pitch.

Second, enterprises have very different types of data. Today’s systems cannot — I repeat — cannot process, federate, and make findable the information in lab notebooks, product data sheets with engineering diagrams, video and other rich media, or the variants of a presentations just changed by the CEO before he or she gives a talk to analysts. Plus, there are other data that will never be in a federated index. Example? How about legal materials germane to a live case.

Third, content processing to make sense of data are available and the successful vendors are not — I repeat — are not companies based on enterprise search and retrieval methods. Examples range from Palantir Technologies to Diffeo, from investigative tools like Shadowdragon to industrial strength systems like BAE NetReveal. These systems include search, but the function is a utility. These software systems are designed to handle high value data analysis. Want a list of documents? Easy. Want analysis? Look elsewhere.

Am I amused with this recent push to make search and retrieval credible? Yes. The scorched earth approach of enterprise search vendors, it appears, has been forgotten, just like the recipe for a break and milk pork chop.

Want substance? Look beyond enterprise search vendors and Fast (no pun intended).

Stephen E Arnold, December 16, 2019

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta