[[!meta title="p-search: a local search engine in Emacs"]]
[[!meta copyright="Copyright © 2024 Zac Romero"]]
[[!inline pages="internal(2024/info/p-search-nav)" raw="yes"]]

<!-- Initially generated with emacsconf-publish-talk-page and then left alone for manual editing -->
<!-- You can manually edit this file to update the abstract, add links, etc. --->


# p-search: a local search engine in Emacs
Zac Romero - <mailto:zacromero@posteo.com>

[[!inline pages="internal(2024/info/p-search-before)" raw="yes"]]

Search is an essential part of any digital work.  Despite this 
importance, most tools don't go beyond simple string/regex matching. 
Oftentimes, a user knows more about what they're looking for: who 
authored the file, how often it's modified, as well as search terms that 
the user is only slightly confident exist.

p-search is a search-engine designed to combine the various prior 
knowledge about the search target, presenting it to the user in a 
systematic way.  In this talk, I will present this package as well as go 
over the fundamentals of inforation retrieval.

Details:

In this talk, I will go over the p-search.  p-search is a search-engine
to assist users in finding things, with a focus on flexibility and
customizablity.

The talk will begin by going over concepts from the field of information
retrieval such as indexing, querying, ranking, and evaluating.  This
will provide the necessary background to describe the workings of
p-search.

Next, an overview of the p-search package and its features will be
given.  p-search utilizes a probabilistic framework to rank documents
according to prior beliefs as to what the file is.  So for example, a
user might know for sure that the file contains a particular string,
might have a strong feeling that it should contain another word, and
things that some other words it may contain.  The user knows the file
extension, the subdirectory, and has known that a particular person
works on this file a lot.  p-search allows the user to express all of
these predicates at once, and ranks documents accordingly.

The talk will then progress to discuss assorted topics concerting the
project, such as design considerations and future directions.

The aim of the talk is to expand the listeners' understanding of search
as well as inspire creativity concerning the possibilities of search
tools.

Code: <https://github.com/zkry/p-search>



[[!inline pages="internal(2024/info/p-search-after)" raw="yes"]]

[[!inline pages="internal(2024/info/p-search-nav)" raw="yes"]]