More about Good Enough Search

October 10, 2016

I have concluded that finding information is entering a mini Dark Ages. The evidence I have gathered suggests that young folks will speak to their mobile devices to get pizza and information for their PhD research projects. I have a folder of examples of applications of smart software which produces remarkable marketing assertions and black box outputs.

I have added to my collection the write up “Postgress Full Text Search Is Good Enough.” I learned that “good enough” search has these features:

  • Stemming
  • Ranking / Boost Support
  • Multiple languages
  • Fuzzy search for misspelling
  • Accent support

I assume, which is risky, that keywords are part of the basic feature set. But in a world of “good enough”, who knows?

The write up provides code snippets for and details regarding the implementation of Postgress’ search function. The explanation of Postgress’ internal methods may require that you keep some Postgress manuals handy and have a browser pointed at Bing or Google to chase down some of the jargon; for instance:

A tsvector value is a sorted list of distinct lexemes which are words that have been normalized to make different variants of the same word look alike. For example, normalization almost always includes folding upper-case letters to lower-case and often involves removal of suffixes (such as ‘s’, ‘es’ or ‘ing’ in English). This allows searches to find variant forms of the same word without tediously entering all the possible variants.

The section on optimization and indexation provides some useful guidelines. Trouble may result from mismatching one’s data with the types of indices Postgress offers.

If you are using Postgress and interested in “good enough” search, you will find the write up helpful. If you are an entrepreneur and want to tap into an underserved market for a graphical administrative interface for Postgress “good enough” search, you will find that the write provides a checklist for you to follow.

For me in rural Kentucky, I marvel at the happy acceptance of “good enough” search. Once “good enough” takes hold, where does one find the impetus to deliver outstanding search? Do I look to dtSearch? IBM OmniFind (aka, Watson). A whizzy cloud service like Amazon’s?

I suppose I can ask Siri or Cortana. Good enough search and good enough answers. Except when the answers are off point or just incorrect. A “C” is good enough for today’s business and technical approaches.

Stephen E Arnold, October 10, 2016

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta