Skip to main content icon/video/no-internet
icon backReturn to Entries

Humor Detection

The term humor detection (also known as computational detection of humor) refers to the computer's ability to detect (but not necessarily understand) some forms of humor. Most, if not all, humor detectors so far have addressed humor within text but not ventured outside of text-based humor. There are two general methods to humor detection. The first is based on some characteristics of text that a machine finds to correlate with humor without using theoretical foundations of humor. The second one is explicitly based on some humor theory that is typically implemented for a restricted subset of humorous texts. This entry discusses general approaches to humor detection and some of the implemented work in the area.

Humor detection is a relatively new area within computational humor, but it is becoming increasingly important with the development of socially intelligent computing, where a human-friendly computational system models human interaction in its communication ability. Because humor plays a significant role in the lives of most people, it has been argued that a computer should understand, or at least detect, humor to be able to function as a companion to a human.

Applications

Suggested applications for computational humor detection vary from detection of unintended humor to reacting appropriately to the intended one. For example, it is easy to imagine an application that would proofread a written speech, flag all instances of potential humor, and let a human revise those instances that were not meant to be there. Because humor is considered to be one of the latest stages of mastering a new language, a humor detection system could be employed in testing people's ability to generate a language-dependent joke in a new language.

A computer plays the role of an intermediary between a person who needs information and a person that recorded it. If a person makes a query for information about a particular product, then the computer must sort through texts to find semantically relevant information. However, many texts contain humor. The humor may be irrelevant to the needed information (in which case humor can be ignored), or the needed information may be directly affected by humor in the text (in which case humor needs to be understood). In both cases, humor needs to be detected for the text to be judged as relevant or irrelevant to the query.

Finally, advanced computational systems are being used in the homes of elderly shut-ins and in hospitals; thus it would be useful for a computer to differentiate a serious request from lighthearted teasing. These applications, however, are sufficiently advanced and above the level of the existing, implemented humor detectors that they are not discussed further in this entry.

Approaches to Humor Detection

There are two theoretical approaches to the computational detection of humor. One relies on foundations of the humor field as a scientific discipline and draws from its theories. The other relies on the observations about some characteristics of a given subtype of humor. In practice, the purely theoretical approaches are hardly used, and even when a detector is built with a humor-based theoretical foundation, some observations about the type of humor that is being detected are used as guided heuristics.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Read next

More like this

Sage Recommends

We found other relevant content for you on other Sage platforms.

close

Have you created a personal profile? Login or create a profile so that you can save clips, playlists and searches

close
close
close Signed In to profile You are signed in as:
close
Logged in Institution Institution