384
You have probably heard the term algorithm before, but many people find it difficult to explain what exactly it is. We show you how to explain the term.
What an algorithm is: This is what lies behind the term
There are different definitions of an algorithm depending on the area of application.
- The term describes a sequence of instructions that can be used to solve a specific problem.
- The sequence of instructions is often found in software and is available in the source code. However, by definition, an algorithm can also be the written instructions to connect the computer correctly. This allows humans and machines to execute different algorithms.
- You will often come across other criteria such as “finite length”. This means that the algorithm must terminate after a finite number of steps, i.e. it must end.
- If you encounter the term “well-defined” in connection with “algorithm”, this means that the instructions are unambiguous, i.e. not ambiguous, and must exist.
- An algorithm – like any other programme code – is written according to a strict syntax.
Known algorithms: You will find these algorithms in everyday life
Wondering what the abstract concept of an algorithm has to do with everyday life, we have some examples of great algorithms that play a role in work and leisure.
- When you drive to work in the morning and use a navigation device, there is often the Dijkstra or shortest path algorithm behind the device. This calculates which route is fastest in the form of a graph and its nodes. Even factors such as traffic jams, gradients and road closures are taken into account here.
- When you enter a term in a search engine such as Google, your results are output via the PageRank algorithm. It checks the result web pages for competence, linking structure and relevance to your search query. The results are then sorted.
- Behind each traffic light is an algorithm that defines the exact switching behaviour.
- If you read something about an algorithm in connection with social media, it means a kind of filter. It can pick out posts on Instagram or TikTok according to content and hashtags and either play them out to a wide audience or only show them to a few people.
Improving algorithm ethics and conscious handling of personal data
Given the growing importance of algorithms, it is crucial to find ways to improve algorithm ethics and promote more conscious handling of personal data.
- Transparency and explainability of algorithms: The workings and decision-making processes of algorithms are often opaque to users. However, efforts should be made to design algorithms in such a way that they are comprehensible and understandable. Companies and organisations should disclose which data sources and criteria are used in the decision-making process of algorithms.
- Diversity and inclusion in algorithm development: Algorithms often reflect biases and preferences of their developers. This can lead to discriminatory or inequitable outcomes. Including diverse perspectives and experiences in algorithm development can minimise potential biases.
- Regulation and governance of algorithms: Governments and regulators should set standards and guidelines for the use of algorithms, especially in areas such as privacy, protection against discrimination and transparency. Companies should be held accountable if their algorithms are proven to be discriminatory or unethical.