undefined | Better HN

0 pointscastell11y ago0 comments

> it has been changed significantly since then in almost every respect, both algorithmically and in the concept.

Does the about page reflect the "old" or the "new" concept? ( http://acuman.us/about.txt )

Can you describe the ACUMANSCRIPT? Is it similar to Prolog or Lisp? Fuzzy string match is this something like Soundex/Double Metaphone? Do you use a knowledge base or ontology (Freebase, Wikidata, etc.)? What programming language is it written (beside your script lang), Node.js/Go/Python/C++ ?

0 comments

1 comments · 1 top-level

seanlevan11y ago

Thank you for expressing interest and checking out ACUMAN.

The "about page" reflects the "old" concept, as stated on this new page: http://acuman.us/readmore.php

The syntax of the ACUMANSCRIPT markup language is actually more similar to XML (or, in the AI field, AIML) due to its rigid structure. It's fundamentally based on pattern matching akin to regex. However, the backend processing which alters ACUMANSCRIPT and processes the individual data is what makes it unique.

The knowledge base used for the Named-entity recognition for the processing of ACUMANSCRIPT is Wikimedia and WolframAlpha, used in conjunction.

It is all processed in the backend using PHP.

j / k navigate · click thread line to collapse