In a text entry system, the idea of predictive refers to the suggestions that the system makes while the user presses the keys.
The first meaning of the term predictive that the Royal Spanish Academy ( RAE ) includes in its dictionary refers to that which predicts or that is useful for that purpose . Predicting, meanwhile, refers to anticipating something that is going to happen .
Predictive , on the other hand, is an adjective that refers to what is linked to prediction : the act and result of predicting or the expression that conveys what is predicted.
The concept of predictive in technology
Nowadays, it is common for the concept of predictive to be used to name the text entry system used by mobile devices (phones, tablets). Thanks to this technology , the user presses a key and the system anticipates what words the person might intend to write.
When programming a predictive text system for an interface such as an online search form, developers can choose different paths. The simplest model of all simply consists of a database with a large number of previously registered words, with which the text string entered by the user is compared.
In this way, we would have a basic system that simply offers the user a small saving of time and energy only in cases where their intention was to search for one of the terms present in the database. Starting from this we could take a step towards a more complex model and, in the long run, more useful for the person. To do this, we should store those text strings that differ from those previously existing in the database.
Search engines tend to be predictive to speed up searches.
Some examples
Let's see what has been said so far using a practical example: we have a database with thousands of words in Spanish; the user types "microsoft" and performs the search; Since that term did not exist in previous records, the system stores it for future use. It should be noted that the predictive nature of this search engine tells us that in the process it will probably have suggested words such as "minerals" or "micro-patronage" to the user, without success.
Now well. In an ideal case, our predictive text system has a new word. However, what would happen if it had a misspelling or if it were a series of letters that made no sense? This without entering into the most dangerous terrain of attempts to execute malicious code through the search engine . This is the point at which more work becomes necessary to evaluate the text strings before storing them permanently.
This process has several phases, and becomes increasingly efficient as the system has more information. For example, on the first day you may not understand that "computer" is an incorrect term, but after layers of review you will be able to store it if you did not have "computer" or discard it and not accept it again; In fact, this experience would help it suggest the correct word even if it does not begin with the same letter.
Predictive and digital writing
Predictive technology aims to make digital writing easier. To make predictions, it uses the most used terms in the dictionary and the user's history and offers the most probable alternatives.
The user, in this way, has the option of accepting a selection and using it, or continuing to press the letter keys one by one until completing the word.
Other uses of the notion
A predictive model , on the other hand, is a data model that uses statistics to predict the return on an investment or the effect of a marketing action. This type of resources is also applied to anticipate the results of the elections.
Predictive maintenance , finally, contemplates the development of various actions to detect possible problems or failures in a machine and thus avoid serious problems in the future. For this, it usually analyzes symptoms or signs that reflect the possible disorder.