Generalization

A generalization of a concept is an extension of the concept to less-specific criteria, in other words, it is the abstraction of data to a smaller scale. Generalization is the act of simplification. It is a foundational element of logic and human reasoning. Generalization posits the existence of a domain or set of elements, as well as one or more common characteristics shared by those elements. As such, it is the essential basis of all valid deductive inference. The process of verification is necessary to determine whether a generalization holds true for any given situation.

The success of automated generalization relies on translation of human knowledge of manual generalization techniques into explicit rules and logic, so that they can be coded in computer language. Very few manual generalization guidelines exist in textbooks, and they are usually vague and incomplete. The concept of generalization has broad application in many related disciplines, sometimes having a specialized context-specific meaning. By generalizing you can consolidate data and portray the most important information.

For any two related concepts, A and B; A is considered a generalization of concept B if and only if:
 * every instance of concept B is also an instance of concept A; and
 * there are instances of concept A which are not instances of concept B.



For example, Plant is a generalization of Tree because every tree is a plant, and there are plants which are not trees (Flowers, for instance).

GIS data in our modern world is highly detailed, and we need generalization in order to understand our GIS data better. GIS data is almost always too detailed for the purpose of our map. For example, if we want to make a map for the state of Utah, we are not going to put all the roads in Utah onto our map. It is important to remember that after generalization has taken place, detail is lost. When dealing with spatial data in a GIS, this simplified data is now less spatially accurate. This loss of detail should be taken into account when taking measurements such as length, perimeter, or area as these measurements will now contain some amount of error. While generalization is limiting in some ways, it should be viewed as a positive aspect of GIS. No GIS project will include all the details of any particular area, so generalization is inevitable. Generalization is necessary to some degree on all maps (or any other GIS project) in order for it to be understandable, and for its theme to stand out.

Hypernym and Hyponym
A hypernym is a term that generalizes a class or group of equally-ranked items, or a word that more specific words would fall under categorically. For example, tree is a hypernym for peach and oak, while ship is the hypernym for cruiser and steamer. Conversely, a hyponym is a term that has a type-of relationship with a hypernym. A hyponym is a word that is more specific than a than a general term that is related to it. Hyponyms such as maple and oak are subcategories of the tree hypernym while cruiser and steamer are types of the hypernym ship. A hypernym is superordinate to a hyponym, and a hyponym is subordinate to a hypernym. This kind of generalization versus specialization (or particularization) is reflected in the mirror of the contrasting hypernym and hyponym word pair. Understanding how to use and apply hypernyms and hyponyms can greatly assist with generalization in cartography.

Generalization in Cartography
Generalization has always played a significant role in cartography. Maps cannot convey every detail of the real world, so cartographers must decide what information is needed to convey their message. Then they generalize their map content to create a product that conveys the desired geospatial information within their representation of the world. Every map has been generalized to some extent and correctly generalized maps are those that emphasize the most important map elements while still representing the world in the most faithful and recognizable way.

Cartographers use maps to tell a story and they have to know their audience, just like a person who is telling a story to their friend. That story needs context and details to make sense, but not too many details so that the listener gets confused. Generalization is used in maps to get rid of the unnecessary details to not confuse the viewer.

Different techniques or operations have been developed to remove, enhance and subdue cartographic visual detail. These operations include content operations, geometry operations and symbology operations. For a detailed explanation of each type of operations, see section on generalization operations in cartographic generalization.

GIS and automated generalization
As GIS came up in the last century and the demand for producing maps automatically increased, automated generalization became an important issue for National Mapping Agencies (NMAs) and other data providers. Thereby automated generalization describes the automated extraction of data (becoming then information) regarding purpose and scale. Different researchers invented conceptual models for automated generalization:
 * Gruenreich model
 * Brassel & Weibel model
 * McMaster & Shea model

Besides these established models, different views on automated generalization have been established. Two such views are the representation-oriented view and the process-oriented view. The first view focuses on the representation of data on different scales, which is related to the field of Multi-Representation Databases (MRDB). Topographic maps and navigational charts are often representation-oriented. The latter view focuses on the process of generalization. This is done by taking a detailed database and reducing its complexity depending on the scale size.

Databases
In the context of creating databases on different scales, two common strategies are the ladder approach and the star approach.


 * Ladder-Approach

The ladder-approach is a step-wise generalization, in which each derived dataset is based on the other database of the next larger scale. For example, if the objective was to thin out details for a map which would be at a scale of 1:250,000, the cartographer would first start at 1:24,000 and generalize details there. She would then move to 1:50,000 performing the same operations. This incremental increase would continue until the cartographer is satisfied with the appearance at the final 1:250,000.


 * Star-Approach

The star-approach describes the derived data on all scales is based on a single (large-scale) database. The cartographer has the same goal in mind as she did in the ladder-approach, but instead of an incremental increase, the cartographer would use the 1:24,000 and go directly to 1:250,000 or 1:100,000 and perform generalizing operations there.