Academic Signature in the search engines


This  report, written on june 26, 2015, gives information on the visibility of the original homepage of Academic Signature (https://www.fh-wedel.de/~an/crypto/Academic_signature_eng.html) in various search engines. I wrote it after a sudden deletion of the webpage from US based google searches.

 

Academic Signature and google:

Google has generally treated academic signature well. From the very beginning, it slowly climbed in page rank. Before the summer of Edward Snowden in 2013, it ranked top only in the neutral search "open source elliptic curve digital signatures".

After the snowden exposures in 2013 it also ranked on the first result pages for google:"open source public key cryptography" or "open source elliptic curve cryptography". This has been true for searches done from european or north american IP-adresses. The googlebot visits academic signature frequently and google has the page and previous versions of it cached.

Acadmic Signature in Google via disconnect.me:
Disconnect.me is a search engine interface, that claims to enhance privacy and allows to select the country of origin for google searches. Info from disconnect.me tells all their servers are located in the US and the UK.

Recently after a five eyes critical posting  in a news forum(june 24, 2015) and referring to a google search "open source elliptic curve cryptography", within hours it practically disappeared from US-based google search results. At least https://search.disconnect.me/ claims that, direct access via an US tor exit node on June 30 2015 showed academic signature ranked 2 on google-US. So it is likely, that google-US is indeed clean and that only "disconnect.me" was manipulated.

At any rate Academic Signature remained ranked among the top 4 in e.g. German, French or even Canadian based searches google:"open source elliptic curve cryptography" even via  https://search.disconnect.me.

This specific "censorship" in the US had happened before already. At that time, the google page rank had recovered within some days.

 

Academic Signature and Microsoft's Bing:

The relationship is simple. Bing had Academic Signature practically blacklisted in 2013. As far as I have tried, this is true for searches  from all parts of the world alike. Before 2013 - with a focus on digital signatures and before mentioning Edward Snowden - it had been highly visible in Bing. Since then you'll never find it with an unspecific search about open source public key crypto, elliptic curve cryptography or the like.

If you force Bing to find it with a named search(search: academic signature cryptography), Bing claims the web page had forbidden to Bing to index its content. This is not true.
I suspect the name Snowden is to Bing like a crucifix to vampires ;-) Their hard disks decay to dust if you try to store the name Snowden on them.

Recently the bing crawler started visiting academic signature's homepage several times a week without any effect on the page rank. Yahoo, owned by microsoft, even accesses project source code or binaries almost on a daily basis, yet Academic Signature's university homepage is kept out of search results.

 

Academic Signature and Yandex:
Before about 2013 you could find Academic Signature easily in Yandex. "open source elliptic curve digital signatures" yielded a hit on the first page. Shortly thereafter, it had been blacklisted. After about 2014 the visibility slowly improved from extremely poor to increasingly better visibility.

As of now (June 2015) on a europe based search (Yandex: "open source elliptic curve cryptography") performed at the time of writing it ranked 15, on the second page of results. I'd like that higher, but this seems quite normal and neither suppressed nor boosted. The yandex spider visits the homepage about 2-3 times a week.

 

Academic Signature and Baidu:

Initially Baidu did not index academic signatures homepage. In the last couple of years, it made it into Baidu's index and the page rank slowly increased. As of today Baidu: "open source elliptic curve cryptography" yields rank 8 on the first result page of Baidu(searched from a german ip-address).

The Baidu crawler visits academic signature home about once a week.

 

Academic Signature and Duckduckgo:

It used to be accessible from  duckduckgo, albeit with a very low rank. As of now the assault found in Google/US coinciding with my critical post seems to also fully apply to duckduckgo. Academic Signature vanished from search results for DuckDuckGo:"open source elliptic curve cryptography".
So I speculate that duckduckgo may have at least as much anticipatory obedience towards government requests as Microsofts Bing. This is in stark contrast with duckduckgo's image. (Of course you are free to assume I just don't like duckduckgo because it considers my site irrelevant ;-)...

 

Academic Signature and Fastbot:

Fastbot is a german search engine maintaining an own independent index. As of today Academic Signature ranks top on fastbot:"open source elliptic curve cryptography" and 14 on "open source public key cryptography". This is not as good as the friendly european google listing, but I perceive it as fair and reasonable.
Update: On September 3rd, I noticed that academic signature was removed from fastbot's index. They don't even show GnuPG on "open source public key cryptography" any more but present numerous completely useless links.

 

Academic Signature and meta search engines:

Meta search engines usually yield an averaged rank regarding the included primary engines. As verified today(june 26 2015), there are unusual results for "open source elliptic curve cryptography" on the german meta engine www.metager.de .

In the default setting, metager: "open source elliptic curve cryptography" on one occasion I found many of my pages with references to academic signature, but not the homepage of academic signature itself. This is a clear indication that is has been selectively suppressed in the results. The positive rank in fastbot or yandex is probably superimposed by a more powerful censorship caused by an unknown entity, which is also included as primary source. It seems that in my test in this meta search engine, specific suppression by an unnamed source is taken into account at least as strongly as positive indexing by others, which are named. 30 Minutes later, the same test revealed an other result and drew information from entirely different primary search engines...
To be honest - to me this stinks, and I recommend to avoid metager altogether. This is in stark contrast with metager's public image.

 

To sum it all up:
Regarding the non commercial homepage of Academic Signature, I see no sign of manipulation in:
(non Us)Google, Russian Yandex and Chinese Baidu.

I suspect manipulation in:
Bing, Duckduckgo, Metager, Disconnect.me(temporarily) and newly in German fastbot.




Update 10/3/15:
As of September 3rd 2015, Academic Signature rose somewhat in page rank on Russian Yandex and Chinese Baidu, is still ranked very high in Google, is still practically blacklisted in Bing and other Microsoft influenced search engines and seems to be newly deleted in the german search engines fastbot.de and metager.de.
If I try to force the German search engines to find Academic Signature with named searches avoiding the keyword "encryption" or "cryptography" but rather use "elliptic curve digital signature", a link to my newly created domain academic-signature.org is given. This alternate page has very few other places linking to it, but avoids mentioning the name Edward Snowden.
Apparently it seems to be detrimental to visibility in the German search engine fastbot.de to mention the name "Edward Snowden".
Additionally, using certain search words seems to "clean" the result list. A link to my crypto tool is now withheld if you search for "cryptography" or "encryption" whereas it is still shown in searches for "digital signatures".  Even the very well known tool GnuPG falls victim to this "cleansing" (fastbot search: open source public key cryptography).