Heuristic filtering refers to the use of various algorithms and resources to examine text or content in specific ways. The word heuristic describes a type of analysis that relies on experience or specific intuitive criteria, rather than simple technical metrics. The use of high-level algorithms allows for heuristic analysis of content, where humans can program computers to think in certain ways rather than just applying a purely quantitative analysis.
Heuristic filtering is most widely used on the Internet to filter email and Web access.
One common type of heuristic filtering is called Bayesian filtering. This is commonly applied to filtering email spam. Bayesian filtering helps a computer to recognize certain words and the likelihood that they're related to spam. In general, techniques like this involve training the machine to apply a higher-level analysis of content in order to filter out spam.
Anyone who has a public email service with a spam folder has probably seen some type of heuristic spam filtering at work. As spam continues to be a major challenge in email communication, technology experts are trying to come up with more diverse ways to identify spam. Alternatives to heuristic spam filtering include IP blacklists and URL filtering, which use IP addresses and other information rather than trying to identify spam through content analysis. Critics of heuristic spam filtering argue that a filter based on word association is not always entirely effective in separating spam from legitimate email.
0 Comments