Overview of CourageousSearch (CS)

CourageousSearch is an opensource project intended to empower Power Users and beyond.
The goal is to augment the current rush to Machine Learning and LLM.
More importantly however: it bravely posits that technology should empower people rather than attempting to replace personal thought and analysis.
Indeed CourageousSearch celebrates and promotes that marvelous grey matter between our ears.

¡Own Your Own Search!

• What you search: Search Engines
• How you search: Boosting/Filtering, tweaking
• Why you search: Contexts/Intents
• Annotate Results: boost/bury/tag/comment/...
• Analyze Results: Shared & Customizable Analysis
• Collaborate: Share as much as you like

Why be Courageous in Search?

Courageous Search takes more initial effort than just hitting one of the mega-corp search engines.
So why bother?

In short: because humans[h] are still better, faster, and smarter than the machines when we have the right tools for the job. At least for the next few years.
AND: after an initial 'getting started' investment of time, effort, and thinking, the search and discovery process will be significantly more rewarding and less taxing.

Use BogoMips to better use.
Let your brain do what it does best: important decisions.
Let technology do what it does best: empower our courageous brains. :-)

This project is specifically focused on the courageous folks who want to leverage technology rather than bow to it.

¡Own your Own Search!

Take the Power back from the mega-corps and tech-bros.
Take ownership of your discovery process and more.
Make tech work for you, rather than you being the raw material feeding the big black boxes.

Curious? Skeptical? Hopeful?
Read on!

What you search

• Brave Search API
  • start with system default API account
  • upgrade to user/specific API account when ready (i.e. request limits)
• Solr deployment/collections
  • e.g. SolrSystem
• various other sources depending on extending/customizing elements of CS
  • Marginalilia (coming soon...?)
    • self hosted
  • Lucidworks Fusion
  • AWS OpenSearch service
  • SERPs...?
  • internal/custom search engines

How you search

• Control sites/sources
  • boost and bury or hide/block various sites
    • rely on crowdsourcing to start...
    • customize your preferences if/when necessary
      • Brave Goggles - very cool functionality from the folks at Brave (currently beta)
        • discover goggles at Brave
        • growing list in this repo...
• Add enhanced result analysis & filtering
  • add custom/personalized boosting
  • filter/boost sites, authors, documents, content
    • based on system/communal preferences initially
    • optionally switch to more bespoke filtering/boosting/analyzing

Why you search: Search Contexts/Intents

• What is a CS Context?
  • Contexts are user-defined grouping of search actions, i.e. search intent or goal
  • CS starts with a shared foundation of common search contexts
  • users are able to add custom search contexts (shared, or private)
• What is the Benefit?
  • Using search Contexts can help the system provide shared and/or personalized boosts, analysis, and other improvements in results.
  • For the truly adventurous: Contexts help improve machine learning and model building
    • both self-hosted AND OconEco-hosted CourageousSearch encourage users to access/export any and all of their data and perform analysis outside CS when desired.

Annotate Results: for Memory and Learning

• Most parts of CS have multiple possible annotations
  • Contexts, Searches, Sources/Sites
  • Results Docs/Content, Content Fragments (e.g. paragraphs)

Analyze Results: Shared & Customizable Analysis

• CS has several analysis "agents" built-in
  • coming soon...
  • e.g.:
    • detect code snippet language(s) (i.e. only Scala, not python examples)
    • detect software versions (e.g. JDK 11+, Excel 2010+, Autocad2018+,...)
    • detect author/publisher
    • detect content date
    • apply machine learning pipelines to result document, augmenting stored results/content (e.g. SparkNLP)
    • auto-tag result docs (e.g. import Firefox bookmarks with tags, CS will use ML to predict tags on new content)
    • ...
• Create/use bespoke analysis agents as desired
  • users can create private or shared custom analysis agents

Collaborate: Share as much as you like

• Common elements to share:
  • Contexts, and related:
    • searches, results, annotations
  • Analysis agents
    • simple regex/term matching/boosting
    • more advanced NLP/NLU matching/boosting
• "Offline" deep analysis
  • Machine learning
  • Model building/deployment

Getting Started

• Request a hosted account via project issue
  • note hosted accounts are currently (Feb 2024) very limited
  • more available hosted accounts will be coming in the near/medium future (driven by demand/interest)
    or
• Go to README.Deployment.md and host your own search
  • Note: the repo is currently several months out of date. Read below...

More information

other & misc

See Readme.Developer.md if you have interest in the code, grails, etc

Note: the repo is currently several months out of date. Sorry :-/
Raise a project issue voting for me to spend a day syncing the base of the current CorpusMinder project back to CourageousSearch if you are interested in deploying your own version.

footnotes etc.

[h] Humans: (us, people, or as some future SkyNet might call us: Meat Puppets)