New way of forum viewing

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
Hello everyone.

Wanted to give you advance notice of a great new way to view in the future to view the content of this site.

Some of you know that I have a programming background. So a few months ago I decided to tackle a problem no one has: how do you only see the posts you want and no other? "Ignore" feature doesn't do it as it filters by person. What you want is a filter by content as if there was a person who was screening the posts for you, telling what you would want to look at, and what to ignore.

As a way of example, a quick search will show you Bill Gates' likely email address. If you send him anything to that address externally, it will be read by a screener, not him. The person reads the messages and if appropriate, would forward the email to someone to address, including Bill himself. Wouldn't it be great to have the same capability when you read the forum? An agent which would scour the whole forum, only bringing you the subjects you would be interested in? And get rid of what you don't like?

With Bill literally getting millions of spam emails a day -- you know, folks reaching out to him with "get rich schemes" :D, Microsoft research (MSR) was working on a tool to automate the process of filtering, reducing the ever increasing human manpower that was necessary to handle all the traffic.

When I left Microsoft, they were not even close to being finished. As Bill used to say, the intelligence of our computers is still well below that of a dog, let alone a human! Having kept good relationship with MSR folks, I thought I ping them to see if they had made any progress. To my amazement, they had! I went and saw a demo and it was remarkable.

You would go through a 20-question interview that would take about 10 to 15 minutes. After that simple training, it was remarkable to see how fast and accurately it would filter Bill's emails. For example, by saying you are already wealthy, it would smartly know to get rid of all spams related to that. But not for example your friend asking to take you out to lunch and offering to pay for it. Existing solutions would be too dumb, and look at the word "pay" and throw out the good with the bad.

Even more interesting for the subject at hand was their willingness to license the code to me. MSR wants to justify their cost to the company and instead of just waiting for internal groups to utilize their research, they are now free to license to anyone. And seems like ex-Microsoft employees get a break in licensing cost :).

So I took a snapshot of the code back in November and have been working on it nights and weekends. Integrating it with forum software took some doing as unlike MSR code, Vbulletin code is not written to the same standard (no jokes guys about Microsoft's ability to write code :D). Last night I managed to get it all working around 2:00am in the morning. It still has a lot of bugs but boy, the potential is incredible. It instantly showed me posts I had missed, and filtered out the "What's new" section to a handful of threads for me to see -- right on the money.

Next is to have Steve, Ron and Lee test it out before I roll it out to everyone. After than internal testing, you will be able to play with it. For now, I was so excited I thought I should share the development with you all. :)
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
It is actually a new way of doing this. For example, it will analyze posting style of members and build a ID that is the same across all of posters post. The best analogy is music identification. There, we analyze the first say, 30 seconds of music, and compute a single "hash." The Hash is an ID that survives all transformations of that music from filtering to compression. That way, you can use your cell phone and have it still capture the ID of the song. Same here but of course, far more sophisticated.

For example, some of you know that I misspell people's names sometimes. The software scans my posts, and realizes that I tend to do that. In that sense then, it is able to track my posts, even if I logged in as someone else!!!

Same idea can be used to tell the system you don't wan to hear about digital topics. This is the area that is a bit buggy but when it works, it is astonishing. For example, it rightly detected a discussion around jitter as being about digital audio reproduction based on similar posts where the words jitter and digital went together. It has built an incredible dictionary already. I think it is around 5 Gigabytes already!
 

NorthStar

Member
Feb 8, 2011
24,305
1,323
435
Vancouver Island, B.C. Canada
Which Bill?

I'm almost two weeks into a seven week cross country (USA) trip and haven't been doing much forum reading and posting. Have I missed anything unusual on this forum?

Bill

Yes, Bill Gates himself! ...Certainly not you! :(:p;)

And you missed all about Life and all it's most intimate secrets! :D
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
Guys..... I have some bad news..... The above was an April Fool's joke!!! Sorry. Someone had to celebrate the occasion :D

Turns out MSR has a group which focuses on such issues and they have done clever work in harnessing information in social groups such as ours. The purpose though, is to provide it as answers to technical questions asked by users. Screening posts would require near zero false positive rate which is not in the cards yet.
 

garylkoh

WBF Technical Expert (Speakers & Audio Equipment)
Sep 6, 2010
5,599
225
1,190
Seattle, WA
www.genesisloudspeakers.com
Damn! Relevance-based search is the Holy Grail of search technology.

The two guys who founded Junglee got bought by Amazon in 1996 when their "relevance engine" was nothing more than a 3-minute elevator pitch at Internet World for $200million. I really thought that Amir would be the next Google.
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
Strangest April Fools joke I've ever experienced.
Geek jokes don't always translate well :D

Let me try this one on you.

I used to have 20 PhDs working for me, focused on figuring out how to compress audio and video better and better without sacrificing fidelity. One of them came to me one day and shows me this file that is 1000 times smaller than the original yet he says, it is identical to the original. So I ask him, "OK, play it for me so that I can see." He says, "well, I have written the encoder, but haven't quite figured out how to build the decoder!"

:D

How was this one?
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing