What qualitatively new appeared in search engines after 2000?

Ivan_korolev2015-07-07 11:20:56

ubuntu

Ivan_korolev, 2015-07-07 11:20:56

The main ideas of the concepts of search engines - qualitative leaps:
83g. Entertainment portal about everything. AOL
94 Sort by catalogs (originally manual labor). Yahoo
95. Sort by keywords. Excite
95g. Multilingual search. AltaVista
96. Finding accurate answers to questions (originally manual labor). Ask Jeeves
97 A herd of experts writes articles on popular topics. ABOUT
97 Sort by rating. Rambler
97. Search taking into account the morphology of the language. Yandex
98. Sort by external links. Google
99. Search spider. Teoma
And also all sorts of other things that I find it difficult to attribute to the historical sequence:
Overture - Direct advertising and arranged an auction for a place in the search
Topix - organized news by date, time, topic and place
Scirus search for scientific information
GigaBlast - fix layout and hints for query specification
FindSounds - search for sound files
Ditto - search for pictures
Dictionary.com - search for words
Youtube - search for
videos Inktomi has improved regional search
Social networks... Mail...
I would like to know in what years an algorithm appeared that gives weight to unique words that gave birth to beancor.
And the so-called "naturalness" in the search engines - is it a quantitative change or a qualitative leap?
Has anything else appeared and when? Or are search engines stagnating in development today?

Answer the question

In order to leave comments, you need to log in

7 answer(s)

just_hank_moody, 2018-11-01
@just_hank_moody

If htop is running at the time of the fall, then apparently the network does not fall, but one of the services is turned off.
In this case, it is better to look at the database and web server logs. Perhaps there will be examples of errors.
As an option, install a monitoring system that monitors the status of each service in order to automatically record which service crashes at different times of the day.

Ergil Osin, 2015-07-07
@Ernillew

Look by years at Nygma and WolframAlpha.
I understand they are from 2000.
Well, plus, algorithm changes are coming, where do they go.

myfirepukan, 2015-07-07
@myfirepukan

From the last: today, nothing fundamentally new has appeared in search engines, in fact it can be called stagnation, you can call it a gradual improvement of the original algorithms.
Then it was just a story of the struggle for the quality of the issuance of the part you already wrote and everything revolved around it,
at first it turned out that the rating just didn’t roll, there were morphology and links,
then it turned out that the BM25 algorithm is not ideal and is easily cheated by spam - we fix it. Today, all this is combined into one algorithm + seasoned with any personalization, etc.
it further it turned out that the links are not ideal and they are also winding up - we are correcting
introduced social factors, but they are not enough and winding up - we are correcting
introduced behavioral factors, but they are also winding up - we are fixing
In my opinion, this is a dead end, and in general the question is whether we need such a general search, maybe it can be divided into separate topics. And in general, social networks rule, all sorts of personal recommendations, cuts.

Mikhail Lyalin, 2015-07-07
@mr_jok

since 2000, advertising delivery, socialization and orientation towards the average user have grown qualitatively

sim3x, 2015-07-07
@sim3x

Wikipedia appeared
everyone implemented wiki-based machine learning

�

⚡ Kotobotov ⚡, 2015-07-07
@angrySCV

Now a culture of real-time data consumption is being formed, which cannot be provided by classical search engines with their common databases, which include everything that is on the Internet.
Services and search engines appear inside data aggregators, providing answers that change in real time.
people want to receive answers to very specific requests related to certain types of services -> for example, "what's new" from VKontakte users right now, what Sobchak wrote (again, right now, and not after 2 updates). well, although, for example, Google makes good topics in the form of assistants like "hum naw", which I work based on the context -> also an interesting direction of development.
Attempts to create fast bots by search engines and speed up their results only lead to a worse overall ranking. . .
so that search engines evolve, but only in a different place, in a different capacity.

Oksana, 2015-07-07
@pushno

if, look globally, then everything