📚 node [[agora-search]]
The agora is in essence equivalent to a distributed wiki.
- Note you could theoretically use a wiki to crowdsource search results; in the limit, think of having a wiki page for every likely user query.
Perhaps agora search tries to find a good match for your query locally and redirects to other agoras in the network or to google or duckduckgo if a good local result isn't available.
- When it volunteers something, it also offers to continue your query in the search authority of your choice.
- push and pull could be used to offer hints for how to route queries to relevant results.
The agora can already be used as a search engine.
To do so, you can use the following URL template in your browser of choice:
- Searching this way behaves the same way as searching using the text box always at the top of this agora.
- As of 2021-02-27, you can also just use open search to add the agora as a search engine; this works at least on firefox. You should be able to see the agora as an available search engine to add in the search box.
The agora will try to serve any relevant nodes for your queries -- useful for when you have notes on the topic you're searching for.
- In every case, the agora will try to make it easy for you to 'click through' and continue your search elsewhere -- in established search engines or the websites of your choice.
- To do so, you can use the following URL template in your browser of choice:
- Agora Search is crowdsourced search for the open internet.
- user story https://twitter.com/flancian/status/1350835941603221504
- keywords https://twitter.com/flancian/status/1352671899583713288
- pull crowdsourced search
- This agora tries to search all agoras, and also the internet as it manifests an intent to cooperate.
In agora server,
/execdoes the following:
- It takes a query string in argument 'q' (sort of standard).
- It optionally tokenizes the query string (unsure if this is the right stage for tokenizing, or tokenization should be done ad-hoc by search providers to allow for domain-specific tweaks).
- It offers the query string or the tokens to search providers, essentially hooks that volunteer actions or results -- perhaps as bids in a distributed system modelled as a market (later).
- It scores the alternatives and either chooses one for the user (based on programmable rules) or presents them as alternatives.
- The default action is 'redirect to the agora node matching/encoding the query'.
- Also supported currently:
📖 stoa (open document) at doc.anagora.org/agora-search
⥅ related node [[crowdsourced-search]] pulled by user
Any entity can have a node in the Agora.
Search queries are entities.
- Search queries can also be decomposed in constituent entities for better modeling using a grammar, but that is not strictly necessary.
- Search queries are entities.
- The Agora is a distributed knowledge graph by the people, for the people.
- From the above it seems to follow that the Agora is a sufficient platform to implement crowdsourced search.
- pull jump agora search
- go https://twitter.com/flancian/status/1345418507916353536
- You can add the Agora as a search engine in your browser: https://anagora.org/jump?q=%s
⥅ related node [[2021 01 17]]
⥅ related node [[2021 02 27]]
⥅ related node [[2021 02 27]]