Th3 Core

Can you talk about the operating costs and/or business model for mojeek?

(Having studied their results, I think a bigger index and fresher results solves a lot of the relevancy issues, their ranking algo is pretty sophisticated)

Looked quite decent at a glance. Couple quick bits of feedback

• One aspect that is tough is some B2B bleed into B2C queries. As an example, Authorize.net embedded in many sites, so is ranking #1 for credit cards. Some of the big search engines with lots of end user clickstream data likely get around this problem by feeding the normalized click data back into rankings & using it to augment or validate the link graph.
• Second thing I saw is lots of listings with no meta description or snippet.
• Third would be not sure I love the brand name. Where did the name come from? Is it synonymous with something else?

(Having studied their results, I think a bigger index and fresher results solves a lot of the relevancy issues, their ranking algo is pretty sophisticated)

If they get enough scale to strike partnerships with some leading sites in popular categories that could help cover some such relevancy issues. Some of the bigger search engines might get worse on this front as more news publications erect paywalls.

The problem with the native Reddit engine is that it only uses its onsite data to display results, but combining that data with actual on page content would make for a great search engine.

If the signal were kept secret it could probably work well, but if they announced it & had significant search marketshare that signal would get hit hard quickly.

>mojeek.com

For a relatively small DB the results I got were pretty good.  Some of the warhorses were missing but everyone displays them because they are copying the Google serps as much as they can. mojeek is ranking for itself which brings up interesting stuff that are just as relevant as the warhorses.

This all makes me wonder if good metasearch engines might make a comeback?  Part of what killed metasearch was all the engines and directories died so it was just Google and Bing and engines using Bing, not very interesting.  But with Gigablast (open source) and mojeek, Yandex and others combined with Bing maybe it's time to revisit the viability of metasearch.  (DDG is really just a blended metasearch anyway.)

In this thread: http://th3core.com/talk/water-coolerextra/new-sighting-'duck-duck'-as-a-verb/  we talk about searx.me an open source meta search engine that is pretty good and hackable.  Something like that might be interesting.  I wish I had the knowledge to install and maintain the script because I'd put it on a server, I guess for no other reason than to say I was back into search again.     Or maybe bribe littleman to put a searx instance up on th3core so we don't get tracked by The Man.

While we are talking, does anyone know anything about Wotbox.com?  It was a spidering engine I wonder if they are still active? 

Also Exalead?

It looks like they spidered for awhile in 2012 and then quit.  Backfill quality sort of results.  So I'll put it in the dead column.

For on-going research on several topics from medical to news, reddit is absolutely my best resource. 

For your regularly visited /r's I suggest bookmarking the /new for descending cron order.

But it can be a firehose.  Let me know if you want TM filters.

>Also, don't forget that you can revert to the "old" text version of subreddits. Handier for scanning.
Is new even usable for you? I cannot stand it as it's completely shite for me.

Also recommended is RES, reddit enhancement suite. Lots of handy shortcuts and tools as a browser addon.