Intuitive seizure

The seized or intuitive striking or predictive (of the English predictive text ), also called the dictionary , is a Technologie conceived in order to simplify the input of text on the telephone keyboards.

Initially Patent ée with the the United States in 1985 as a method of telephone call with the deaf , the intuitive seizure knew only applications limited before the advent of the Mobile telephony and sound service of textual messages, where it finds today its utility principal.

But it could in the future know new fields of application with the evolution of the mobile phones in personal assistants and the development of the Onboard computing on board equipment not having keyboard adapted (in the field Automobile for example).

The market

The intuitive seizure such as it nowadays is known is developed and marketed by various companies whose products are in Concurrence.

Most widespread of them was developed by Tegic Communications (become a Filiale of AOL since its repurchase in December 1999): it is about T9, Sigle of “text one 9 keys” (literally “texto on 9 keys”). It is that which is integrated into the apparatuses of the majority of the manufacturers (in particular LG, Samsung, Nokia, Siemens, Sony Ericsson). The notoriety of this mark became besides such as it indicates in the language running the general concept of intuitive seizure.

However one also finds:

the iTap: created by Motorola for its telephones, it is the most serious candidate of T9;
the eZiTap and the eZiText of ZiCorp;
LetterWise, WordWise and EQx d' Eatoni Ergonomics;
SureType, developed by Research In Motion (RIM) for its Blackberry;
AdapTex, of KeyPoint Technologies (KPT);
Q9, method allowing the encoding Sinogramme S used in the writing of the Chinese languages. Developed by the group hongkongais Qcode Information Technology, it does not maintain any relationship with T9, in spite of the proximity of their names.

Principle of operation

With and without the intuitive seizure

The keys of a telephone keyboard are generally laid out as on the image opposite: to each key a group of several letters (3 or 4) corresponds. In the beginning, their presence did not have the role to allow the input of text: it is there still the popularization of the SMS which resulted in using them for this purpose.

With the first generations of mobile terminals and their system called MultiTap (or sometimes more simply ABC ), the seizure of each letter requires to press $p$ time the key on which it is registered, where $p$ indicates its position in the group. For example the letter Y figure in position $p$ = 3 of the WXYZ group registers on the key 9: it is thus necessary to press the latter with three recoveries to post a Y.

Let us suppose that a user wishes to type the Mot SAGE. He will have to then press: i.e. 77772433, is in all 8 pressures .

The system of intuitive seizure makes it possible as for him to be freed from these multiple strike: only one pressure per letter is enough. Thus for the same word, a telephone which is equipped with it will make it possible to only press: i.e. 7243, is only 4 pressures on the whole.

Need for proceeding by Conjecture

But of course, where a combination of keys corresponds, in the traditional system, with a word and only one (bijectivity), the system of intuitive seizure must face a problem: the plurality of the combinations of letters corresponding to the same seizure.

One can calculate the number of it besides: for $n$ pressed keys, the number of combinations of letters corresponding is of $\ prod_ {i=1} ^n x_i$ , where $x_i$ is the number of letters registered on the $i$ -ème touches combination ( $x_i$ = 3 or 4 according to the keys).

For example with our combination of $n$ = 4 keys, 7243:

the key 7 comprises $x_1$ = 4 letters (PQRS) ,
the key 2 comprises $x_2$ = 3 letters (ABC) ,
the key 4 comprises $x_3$ = 3 letters (GHI) ,
the key 3 comprises $x_4$ = 3 letters (DEF) .

There is thus

x_1

x_2

x_3

x_4

= 4 × 3 × 3 × 3 = 108 combinations of possible letters. But only one harmony with the wish of the user (here SAGE): the system of intuitive seizure thus will have to resort to the interpretation, in order to determine which.

Use of a dictionary

Obviously, among all these combinations, the majority does not represent any existing word in the language of the user. In fact, in addition to the word SAGE, the continuations RBHF, PCGE or QAGF also correspond to 7243.

The system of intuitive seizure must thus start by eliminating these continuations from inintelligibles letters: it rests for that on a Dictionnaire integrated into the memory of the apparatus, which reference the Mot S most commonly used in the Langue of the user.

Choice of the user

But even by having recourse to a dictionary, a seizure can remain ambiguous. Thus in our example, in addition to the word which the user has at the head (SAGE), there are some in the French language still at least six others which correspond to 7243: PAGE, PAIE, RAGE, RAID, RAIE, SCIE.

Consequently, with each time the user presses a key, a algorithm establishes starting from the dictionary the list of the words which could correspond to the seized combination, then determines for each one of them the probability of being that which the user has with the spirit, and finally updates posting by selecting that which it estimates to be most probably waited by the user.

This last can then:

is to continue the seizure of the word in progress, if the full number of letters is not reached,
is to confirm the proposal if it is appropriate to him (generally by a pressure of the key 0 which inserts a space, however the key allotted to space can vary from one apparatus to another), and to begin the seizure of the following word,
is to traverse, by means of a key known as next (“according to”), the list of the other words corresponding to the seized combination.

Tegic put on line some plays making it possible to learn how to use its product: T9.

Emergence of a new Linguistic concept “”

In a fond of delicacies world of Neologism S, the various words corresponding, within a language, with the same combination of keys from now on are qualified by the english-speaking of textonyms (or txtonyms ), of adaptonyms , cellodromes , or even of T9-agrams . The development team of Tegic itself baptized T9-o the shell S caused by a lapse of memory of the key next .

Thus, to take again our example, it will be said that “wise” is a textonym of “saw”, or that “wise” and “saw” are textonyms .

Among the thousands of others textonyms existing in French, one can quote:

Such examples illustrate the importance to read again its messages well before sending them, in order to prevent that one of the words is not replaced inadvertently by one of its textonyms .

The www.textonym.com site makes it possible to find (in English) all the textonyms associated with a mot.

Effectiveness and optimization

In the example presented (SAGE), the intuitive seizure makes it possible to reduce the number of strike 50 % (one passes from 8 to 4); but the system is all the more effective as the words are long and complex.

In order to increase this effectiveness further, it is sometimes possible to personalize the dictionary by enriching it by words which are unknown for him, in particular the proper names. Certain alternatives of the system also have adaptabilities to the practices of the user (training of the words most frequently employed).

The last versions of the dictionaries take into account the Abréviation S recently generalized by the explosion of the SMS and specific sound language. Thus, although it is not a word, JTM (for “I love you”) will be proposed with the user who will have typed 586.

Autocomplétion

One of the principal differences between T9 and the iTap is that this last presents functionalities of autocomplétion, i.e. it is able to guess the word, the expression, even the whole Phrase wished by the user before even as it did not have to seize the full number of letters necessary.

It is based for that on a dictionary including/understanding of the expressions and the sentences usually used, in order to be able to exploit the Contexte word seized in its manner of determining what is awaited.

Although this device can appear very useful, the iTap is however famous more difficult to learn than T9, precisely because of this additional aptitude. Indeed, as the iTap will seek to supplement by others character S the chain already seized, if the word which he proposes is not adapted, the list of the words to make ravel to reach that desired risk to be longer than with T9.

Random links: