Stop words are frequently occurring function words that some text processing methods filter out or give less weight to. Typical examples are articles, prepositions, conjunctions, and certain pronouns. However, modern search engines do not categorically ignore these words: If a stop word alters the meaning or search intent , it can be crucial for processing a search query.
What are stop words?
The English term " stop words" refers to words that occur very frequently in a language and often convey little thematic information on their own. In German, examples include "der", "die", "das", "und", "oder", "in", "von", and "zu".
Many stop words are function words . Function words establish relationships between other words or fulfill a grammatical function:
- Articles: the, a, an
- Prepositions: on, at, off, at, in, with, after, from, to
- Conjunctions: but that, or, and, because, if
- Pronoun: I, you, he, she, we, my, our
- Auxiliary verbs and common verb forms: am, is, are, was, will, have
Which words are treated as stop words depends on the language, the software, the text being analyzed, and the specific use case. Therefore, a universally valid German list of stop words does not exist.
How does stopword removal work in NLP?
Stopword removal is a possible processing step in Natural Language Processing ( NLP) . Natural Language Processing, or NLP for short, is the machine processing, analysis, and generation of human language in text and speech data. Simply put: NLP helps software to... Click to learn more . A system first breaks down a text into individual units called tokens. Then, previously defined NLP stopwords can be removed. This reduces the number of tokens to be processed, allowing certain analyses to focus more on content-bearing terms.
A classic stopword removal can be useful, for example, in simple frequency analyses, keyword evaluations, or certain document search methods. From "the new website for the craft business," the terms "new," "website," and "craft business" might remain after a simple filter.
Removing stop words doesn't automatically improve the result. In tasks that aim to capture sentence meaning, relationships, tonality , and intentions, function words can provide important information. Therefore, when optimizing NLP, you must decide, based on the goal and the available data, whether stop words should be removed, weighted less, or fully considered.
Punctuation marks are not stop words.
Special characters and punctuation marks should not be equated with stop words. Periods, commas, hyphens, and question marks are characters or tokens in their own right. Some processing systems also filter these characters, but that is a separate technical step.
Do modern search engines ignore stop words?
No, modern search engines don't generally ignore stop words. Search systems process the words in a query within their context. Even short and frequent words can be crucial if they alter the meaning.
In 2019, when Google introduced BERT to search, it explained that the models used process words in relation to the other words in a query. Google specifically emphasized that prepositions like "for" and "to" can be important for understanding complex or naturally phrased search queries. This is described in the official Google article "Understanding searches better than ever before".
Not every function word receives the same weight in every query. A search system can give less weight to frequently used words and still consider them if they determine the context, a relationship, or the desired result.
How do stop words change search intent?
An example from South Tyrolean tourism illustrates the difference:
- “Flights from Bolzano” designates departures with Bolzano as the point of origin.
- “Flights to Bolzano” refers to connections with Bolzano as the destination.
The prepositions "from" and "to" appear in many classic stop word lists. However, in these search queries, the prepositions determine the direction of travel and thus the desired information. If a system were to remove both words across the board, only "flights to Bolzano" would remain, and the intent would no longer be clear.
I regularly encounter such differences in my work with South Tyrolean SMEs. Especially with tourist, local, and multilingual offerings, short words can alter the location, direction, or purpose of an inquiry. Content encompasses all intentionally published digital content on websites, in online shops, on social media channels, in newsletters, and in other digital environments. Therefore, "If you want to know..." and " Click and learn more " must not only contain key terms but also clearly answer the specific question a person is asking.
Should you remove stop words for SEO purposes?
Removing stop words is not a standalone SEO measure. Write texts, page titles, and headings naturally and unambiguously. Artificially removing articles, conjunctions, or prepositions often results in difficult-to-read formulations with unclear meanings.
For stop words in SEO: SEO explained simply: SEO is the strategic optimization of a website for organic visibility, relevant search intent, and clear presentation in search engines and search-related answer systems. The goal... Click to learn more.
- Don't remove function words from your content across the board.
- Formulate headings and page titles in an understandable way, instead of following supposed machine rules.
- Keep prepositions when they indicate location, direction, target audienceDefinition of the target group A target group (also target group, target audience) is a specific group of people or buyer groups (such as consumers, potential customers, decision-makers, etc.)... Click to learn more or clarify the intention.
- Evaluate keyword phrases in the context of real search queries and not just as a collection of individual nouns.
- Use a stop word list only for a clearly defined technical or analytical purpose.
A sound SEO strategy aligns content with the needs, context, and language of the target audience. Naturally written content retains its meaning and context, which can be lost if frequently used words are simply removed.
Example of a German stop word list
The following selection shows common words that can appear as potential stop words in German NLP applications. This list is an example and not a universally applicable standard. Depending on the software, corpus, and task, words may be added, retained, or weighted differently.
Articles, pronouns and related forms
- the, the, the, the, of the
- a, an, a, a, a
- I, you, he, she, we, you (plural)
- him, himself, himself, us
- my, my, my, my, my, my, my
- our, our, our, our, our, ours
- his, his, his, his, his, his
- no, no, to no one, none, no
Prepositions and conjunctions
- am, an, auf, aus, bei, bis
- through, for, in, with
- after, without, in order to, from, of, before
- to, to, for
Conjunctions and frequent adverbs
- but, when, so, that, yet
- whether, or, but, and, because, if
- also, there, here, to, still, now
- just like that, on and on, again, really
Common verb forms and other words
- am, are, is, are, was, were, were
- have, has, had, become, will, will
- can, can, could, must, should, ought, want
- all, other, something, completely, every, some, more, nothing, even
- what, which, which, which, which, how
For a technical application, you shouldn't simply adopt such a list without checking it. First, examine what information is lost through filtering and whether the processing step improves the desired result.
FAQ about stop words
What are stop words?
Stop words are common function words that some text processing systems filter out or give less weight to. These include, for example, articles, prepositions, conjunctions, and certain pronouns.
Does Google ignore stop words?
Google doesn't categorically ignore stop words. If a short word alters the context or search intent—sounds dry, right? But in the digital world, it's a real game-changer. Imagine you're searching for "best pizza in..." (Click to learn more ), Google can take the word into account when interpreting the search query.
Should you remove stop words from SEO texts?
No, don't systematically remove stop words from SEO texts. Naturally worded and clearly understandable content is more helpful for both people and search engines than artificially shortened sentences.
Do stop words belong in keywords?
Stop words belong in a keyword phrase when they clarify the meaning of the search query. "Flights from Bolzano" and "Flights to Bolzano" serve different purposes, even though only the preposition is changed.
Should you remove stop words from URLs?
You can remove unnecessary function words from a URL. A URL, short for Uniform Resource Locator, is an address that specifies where a resource is located and the method used to access it. Click to learn more and omit these words if the address remains shorter and unambiguous. Do not automatically remove them if doing so would result in a loss of readability or meaning.
Are stop words allowed in page titles?
Yes, page titles can contain stop words if they create a natural and precise statement. A clear title conveys the topic better than a string of isolated SEO keywords.
What distinguishes stopword removal in NLP from a search engine?
In NLP, stopword removal is a configurable processing step for a specific task and a specific dataset. A modern search engine, on the other hand, processes the search query in context to recognize meaning, context, and search intent.
Is there a legally binding German list of stop words?
No, there is no universally binding list of German stop words. The selection depends on the language, software, text type, and application goal, and should be checked before use.
