Search engines: how algorithms, search operators & reverse search work.

"I'll Google that for a minute." It feels like any question can be answered with the help of search engines. It's not for nothing that the verb 'googling' made it into the Duden dictionary in 2004. But as simple as it seems, it is not necessarily so. To achieve a good search result, so-called search operators help when researching on the Internet. Also the backward search - in the image search - helps. We will show you how to use the reverse image search, what algorithms have to do with the search and much more.

Table of Contents:

1. How a search engine works
2. What are algorithms?
3. Three steps for a good search result
4. 44 Google alternatives
5. What is the image reverse search?
6. How does the image reverse search work?
7. Sources

When we ask a search engine a question - whether single words or whole sentences - it filters the Internet for us. This happens according to a - top secret - algorithm. In fact, the search engine does not filter the Internet live, but rather a kind of "copy" of the entire Internet. Before that, it has "copied" the Internet, so to speak. This "copy" is called an index. A live search would take days and let's be honest... nobody wants to wait that long.

So the so-called search engine index does not show us real-time results, but rather the results at a point in time in the past. Usually this is in the very near past, but it is important to know that the results are not live.

This index is created by so-called crawlers, small computer programs that constantly scan the World Wide Web for new content, capture it and store it in the search engine index.

The "rules" behind all these processes are called algorithms. Based on fixed rules, a series of commands is executed one after the other to solve a task. In practical terms, this means that when you search the Internet, the respective search engine algorithm selects and sorts the results. The algorithm only uses what it is provided with and executes what it was programmed to do. The algorithms therefore determine which results are displayed to you.

The respective algorithms are not only continuously adjusted (which can lead to different results for the same search on different days), but are often not made public by the providers. For example, how the Google algorithm finds and ranks its results is secret. There are organizations that want to make the secrets of the algorithms more transparent, e.g. AlgoTransparency or AlgorithmWatch. These organizations try e.g. to show how the algorithm works on TikTok. You can find more about this in the following links.

1. Use metasearch engines Metasearch engines collect the results of several other search engines and thus facilitate research. Metager is the best known German search engine. It forwards the search queries to about 50 other search engines.
2. Use search operators To search effectively on the Internet, you can also use so-called search operators. These are characters or combinations of characters. The search word is completed, for example, with AND or OR. If you put search terms in quotation marks, e.g. "AI in the world of work" - then you will only be shown results in which exactly this sequence of words occurs. To achieve good results, search commands such as a minus sign also help you to refine your results. For example, you can exclude terms that should not be displayed, such as AI robot.
3. Find multiple matching keywords The search engine doesn't care about correct grammar, it's much more important to find multiple matching keywords. Or enter a short question, e.g. "Where can I find search engines about AI?"

AOL - search results from Bing Baidu - Chinese search engine, censors search results, stores user data, English language search results from Bing BASE - Bielefeld Academic Search Engine: searches more than 300 million documents from more than 10.000 data providers Bing - Second largest search engine in USA & Europe, founded by Microsoft Blinde Kuh - Heavily filtered search results for children, operated by a non-profit association Boardreader - special forum search engine Brave Search - no tracking, no advertising, default search engine of the browser of the same name Brave Disconnect Search - search results of DuckDuckGo & Bing, very high data protection Dogpile -Summary of search results from more popular search engines (Google, Bing, Yahoo & Yandex) DuckDuckGo - high privacy, combination of meta search & own web crawler Ecosia -majority of ad revenue donated to conservation projects, servers run carbon-neutrally, search results from Bing Ekoru - ad revenue goes to conservation organizations that clean up oceans; results from Yahoo (Bing) Exalead -search can be adjusted to specific page types Fastbot - without advertising Fireball - original rock in Germany, high user numbers in 2000s, specializes in German-language search results Fragfinn - heavily filtered search results for children Gexsi -generates donations for social projects, search results from Bing Gibiru - unfiltered private search, search history is deleted seconds after search Gigablast - open source search engine Helles Köpfchen -heavily filtered search results for children, highlights content with high educational value LexiQuo - meta search engine Lukol - search results from Google, but without tracking & data storage Lycos -search results from Bing, formerly independent search engine with its own web crawler MetaCrawler - the "granddaddy of all meta search engines", went online in 1996 as the first meta search engine worldwide shortly before MetaGer MetaGer - meta search engine, went online as the 2nd search engine after MetaCrawler. MetaJob - meta search engine for jobs in Germany Mojeek - English search engine, no personalized data Nona - search engine from Germany, lt. own statement private and (for a small fee) ad-free search engine Oscobo - search results from Bing Paperball - current news from daily and weekly newspapers Qwant - high data protection: no tracking, no advertising, no personalization Qwant Junior - no advertising, highlights content with high educational value Search.ch - search engine for Switzerland & surroundings, numerous special functions Startpage - proxy search engine, search results from Google, but without tracking & data storage Suchmaschinen-Datenbank.de - search engine for search engines Swisscows - Swiss search engine, without tracking, family-friendly due to youth protection filter tiger.ch - own index with webcrawler for Switzerland, plus metasearch (Google, Bing, DuckDuckGo), as well as div. video portals WolframAlpha - delivers direct answers & facts instead of "only" search results WorldWideScience.org - finds more than 200 million documents from about ten countries, including Germany YaCy - best-known P2P search engine, users need pre-installed software, open to everyone Yahoo - once market leader, search results from Bing Yandex - Russian-language search engine, market leader in Russia Yep - according to its own statement, 90% of advertising revenue goes to content creators to build a fair Internet You.com - search results are grouped into several categories to display more search results, high privacy protection

In order to recognize whether certain images also match a certain message, there is the so-called reverse image search. Photos can be used as search queries and entered into the search engine. In this way, it can be clarified, for example, whether a certain image has already appeared in a completely different message than the current one. An important note: Sometimes images are deliberately used in reverse in order to deceive. In order to verify images, it is advisable to use different search engines, as they draw on different databases and algorithms.

For the search you need the image as a file or the image URL (Internet address of the image). You can find the URL of an image by right-clicking on the image. Copy it to your clipboard. The reverse search of the respective search engine can be found in the image search. Click on the camera icon at the end of the search slot. You can then paste the corresponding image URL from the Internet. By the way, the reverse search is also possible with published photos from social networks. The result of the reverse image search will show you all websites with matching and similar photos.

The reverse image search is especially important for detecting fake news and false reports.

BR: https://www.br.de/nachrichten/netzwelt/wie-tiktok-seine-nutzer-fesselt NDR: https://www.ndr.de/ratgeber/medienkompetenz/Wie-wirken-Algorithmen-Unterrichtsmaterial-fuer-die-Schule,algorithmus100.html ReverseImageSearch: https://www.reverseimagesearch.org/de Wikipedia: https://de.wikipedia.org/wiki/Liste_von_Websuchmaschinen YouTube: https: //www.youtube.com/watch?v=2UvK-8DyXyc