Predictive Multitap Text Entry System For Mobiles English Language Essay

Introduction

Text entry has been done in computing machine applications unequivocally utilizing individual keys or sometimes cardinal combinations with 102-key keyboard. However, the upsurge use of Mobile telephones which have a minimum 12-button keyboard prompted the demand to come in text in a most efficient manner. Text entry is by no agencies new in nomadic computer science ; there has been a explosion of research on the subject in recent old ages. There are several grounds for this heightened involvement: foremost, nomadic computer science is on the rise and has spawned new application spheres such as wearable computer science, bipartisan paging, and nomadic web and email entree. Second, word processors, spreadsheets, personal schedulers, and other traditional desktop applications are progressively available on nomadic platforms. Third, there is a strong demand for the input of text or alphameric information that is easy and expeditiously entered, recognized, stored, forwarded, or searched, via traditional package techniques. Fourth, the phenomenal success of text messaging with nomadic phone users has inspired considerable guess on future by-product engineerings, all expected to profit from text entry. The statistics for text messaging on nomadic phones are singular. In January 2001, GSM Europe reported that 15 billion ( 15,000,000,000 ) SMS text messages are transmitted per month worldwide ( Alsio G. & A ; Goldstein M. ( 2000 ) . This is peculiarly interesting in position of the limited capableness for text input with the current coevals of nomadic phone engineering. While the omnipresent QWERTY keyboard reigns supreme as the primary text entry device on desktop systems, nomadic phones and hand-held systems use tantamount techniques for the text entry. But how efficient and simple are these techniques? And so, the challenge of text entry for nomadic phones nowadayss itself. Though the QWERTY keyboard has the obvious advantage of acquaintance, it is bulky and unless the keyboard is life-size, touch typewriting is hampered or impossible. Basically, there are two viing paradigms for nomadic text input: pen-based input and keyboard-based input. User experience with typewriting and handwriting greatly influences outlooks for text entry on nomadic devices.

To maximally use the limited computer keyboards on nomadic communicating devises, Multitap method is frequently used for text entry. The drawbacks of this method include pressing a key more than one time to come in each coveted missive and besides require a important sum of ocular seeking to happen a needful missive on a cardinal doing it comparatively inefficient from the point of view of the figure of key strokes required to come in each word ( Jun, Bryan, & A ; Peter ) .

To get the better of the inefficiencies of Multitap method, Predictive text entry system was developed. Predictive text entry reduces the figure of key strokes when entry texts by analysing the sequence of key strokes being entered and thereby predicts the intended word for the users. However, surveies shows that users are dissatisfied with how bing systems for Predictive text entry work because it interferes with their communicating ( YH af Segerstad, 2003 ) . Multi-tap text entry system combined with Predictive text entry called Predictive multi-tap text entry system works to extinguish this dissatisfaction.

In order to efficaciously predict words, linguistic communication theoretical accounts are used to choose and decode words that correspond to come in key strokes sequence. Language patterning being used includes n-gram and bi-gram word theoretical account, principal and dictionary theoretical account. A proposed theoretical account in this paper is used to implement prognostic multi-tap text entry system for nomadic communicating devices.

Literature Review

Text entry methods could be subdivided into two basic categorizations ; Stylus ( Pen ) -based text entry and Key-based text entry.

Stylus-based text entry uses a pointing device, typically a pen ( or stylus ) , to choose characters through tapping or gesture. Examples of this method include:

Handwriting acknowledgment: This was one time touted as the solution for nomadic text entry, but the system has two major jobs that handwriting recognizers must work out ; cleavage and acknowledgment. However, there are no nomadic consumer merchandises in the market today where natural script acknowledgment is the exclusive text input method. The merchandises that support stylus-based text input work with restraints or stylized alphabets, connoting that handwriting acknowledgment does non execute adequately.

Gesture-based Text Input: The text entry methods in this subdivision are classified as gesture because of their informality and fluidness. Character recognition-based and soft keyboard-based input techniques have fixed characters that are entered in a certain manner, or the stylus must be tapped in a certain location to choose characters for input. Gesture-based text input engineerings do non hold a fixed set of shots that a recognizer turns into characters ; gesture text input methods have a model in which informal stylus gestures are interpreted as characters. An illustration of this is Cirrin, a engineering presented by Mankoff and Abowd ( 1998 ) .

Soft Keyboards: A soft keyboard is a keyboard implemented on a show with constitutional digitising engineering. Text entry is performed by tapping on keys with a stylus or finger. The advantages of soft keyboards include simpleness, and efficient usage of infinite. When no text entry is happening, the soft keyboard disappears, therefore liberating screen infinite for other intents.

The major job with the soft keyboard is the show of nomadic phones. Mobile phones normally have a little show. Therefore, implementing a keyboard on the show means cut downing the size of points on the screen, which could do the show hard to read, or the show could go gawky.

Key-based text entry techniques range from those that use a keyboard where each key represents one or more letters, to those with every bit few as three keys. Examples of this method include:

Small QWERTY Keyboards: The most prevailing text input engineering for low-end PDAs is the illumination QWERTY keyboard ; an illustration is the Nokia Communicator shown in Figure 1. The Nokia Communicator is a nomadic phone with text messaging functionality.

Figure 1: Nokia Communicator 9110 ( existent size 158 A- 112 millimeter )

Joystick Text Entry: This method is similar to the soft keyboard text entry technique. A keyboard ( alphabetic order or QWERTY agreement ) is presented on the show, and the control stick is used to travel the pointer through the alphabet. The coveted missive is selected by pressing the control stick or an ENTER key. This technique is sometimes called the day of the month cast method because, similar to a day of the month cast, the coveted character is selected by revolving through the character set. Video arcade games frequently use this technique for participants to come in their name when they achieve a high mark. The technique is besides normally used for come ining text into some electronic musical instruments. Although this method is sensible for come ining little sums of text into devices with a simple interface, the method is frustratingly slow and non suited for even modest sums of text entry.

Telephone Keypad: The desire for an effectual text entry utilizing the telephone computer keyboard is fuelled by the addition in text messaging services, and the motion toward consolidation of engineerings such as radio telephone and hand-held computing machines. Text entry on most nomadic phones is based on the standard 12-key telephone computer keyboard as shown in Figure 2.

Figure 2: The standard 12 cardinal telephone computer keyboard

The 12-key computer keyboard consists of figure keys 0-9 and two extra keys “ * ” and “ # ” . Fictional characters A to Z are spread over keys 2-9 in alphabetical order. The arrangement of characters is based on an international criterion ( Grover, King, & A ; Kuschler, 1998 ) . Since there are fewer keys than the 26 needed for the characters A-Z, three or four characters are grouped on each key, and so, ambiguity arises. There are three chief attacks for get the better ofing this ambiguity: Multitap, two-key, and one-key with disambiguation.

Multitap: The Multitap method is presently the most common text and simplest input method for nomadic phones. With this attack, the user imperativenesss each cardinal one or more times to stipulate the input character ( Jun, Bryan, & A ; Peter ) . For illustration, the 2 key is pressed one time for the character A, twice for B, and three times for C. A job arises when the user attempts to come in two letters from same key consecutively. To get the better of this, MultiTap employs a time-out on the cardinal imperativenesss, normally 1-2 seconds, such that no cardinal imperativenesss during the timeout indicates completion of the current missive. Although Multitap eliminates ambiguity, it is rather slow, with key strokes per character ( KSPC ) rate of about 2.03 ( Wigdor and Balakrishnan, 2003 ) . For illustration, to come in the word “ ON ” the user presses the 6 key three times, delaies for the system to timeout, and so presses the 6 cardinal twice more to come in the N.

In the two-key method, the user imperativenesss two keys in turn to stipulate a character. The first key selects the group of characters ; the 2nd key specifies the place within the group. For illustration to come in the character ‘K ‘ the user imperativenesss ‘5 ‘ key to choose the group ‘J ‘ , ‘K ‘ , or ‘L ‘ followed by 2 to choose ‘K ‘ which is the 2nd character in the group. It has a KSPC of 2, since all letters require two back-to-back cardinal imperativenesss ( Wigdor and Balakrishnan 2003 ) .

In one-key with disambiguation, the user presses the sequence of keys to organize the needed text, the system so computes all possible combinations of the sequence, look them up in a dictionary and so presents the lone valid combinations from where the users pick the needful 1. This method is besides referred to as lingual disambiguation. Obviously, the term “ one-key ” in “ one-key with disambiguation ” is an simplism! T9 was the first disambiguating engineering to work with a criterion Mobile phone computer keyboard, but non the lone such engineering.

Predictive Text Input

Predictive texting is intended to simplify text entry and to cut down the input load by foretelling what the user is come ining ( Ling, 2008 ) . This can be accomplished by analysing a big aggregation of paperss called principal to set up the comparative frequence of characters, digraphs, that is, braces of characters, trigrams, words, or phrases in the linguistic communication of involvement. These statistical belongingss are used to propose or foretell letters or words as text is entered. Predictive text input method, alternatively of utilizing a individual key for a group of character as it is used in Multitap method, allows users to come in groups of characters per key stroke and allow the constitutional algorithms decipher the input cardinal sequences into most likely intended words and phrases ( Rick et Al ) . Predictive input combined with Multitap ( e.g. eZiTap ) increases the rate at which text is entered with less lights-outs, nevertheless lingual cognition must be added to the system in order to avoid meaningless words. Naturally, lingual disambiguation is non perfect, since multiple words may hold the same cardinal sequence. For illustration, pressing keys 843 will gives a possibility of 27 words, but with lingual cognition, two are valid words are available to be chosen. In this instance, the user must press extra keys to obtain the coveted word. Predictive text input techniques strive to cut down the input load by foretelling what the user is come ining. This can be accomplished by analysing a principal ( big aggregation of paperss ) to set up the comparative frequence of characters, digraphs ( braces of characters ) , trigrams, words, or phrases in the linguistic communication of involvement. These statistical belongingss are used to propose or foretell letters or words as text is entered.

To actuate this attack, see a user composing the message “ Meet at place subsequently ” . Assuming the criterion telephone computer keyboard utilizing T9, a user who has typed the first two words and is identifying the sequence for “ place ” ( ‘4663 ‘ ) in will be shown “ I ” , “ in ” , “ hostel ” , “ good ” as the keys are typed, with “ place ” eventually entered after a ‘Next ‘ cardinal imperativeness. However, a principal analysis can uncover that “ place ” is the most likely word being typed, even when merely three letters have been typed, based on the old word, “ at ” .

Language Modeling

Harmonizing to Yijue How and Min-Yen Kan ( 2005 ) , in linguistic communication mold footings, the n-gram theoretical account with a bi-gram word theoretical account is used to do a anticipation. That is, we select the word with the highest conditional chance given the joint grounds from the typed sequence and old word. Counts are once more estimated from the collected principal. All word suggestions can therefore be ordered by their chance. However, harmonizing to R.Shriram et Al ( 2006 ) , there are a few cautions to see in establishing a linguistic communication theoretical account on a standard principal, these include:

The principal may non be representative of the user linguistic communication: The thought that a principal is “ representative of a linguistic communication ” is questionable. This is because users typically use a much richer set of characters and words than appear in any principal, and the statistical belongingss in the user ‘s set may differ from those in the principal. A simple illustration is the infinite key, which is the most common character in English text ( Soukoreff R. W. & A ; MacKenzie I. S. ( 1995 ) .

The principal ignores the redacting procedure: A principal contains no information about the redaction procedure, and we feel this is an unfortunate skip. Users are fallible and the creative activity of a text message – or interaction with a system on a larger scale – involves much more than the perfect additive input of alphameric symbols. The input procedure is truly the redaction procedure.

The principal does non capture input modes: Text paperss do non reflect how they were created. For illustration, a principal includes both capital and small letter characters. In simple linguistic communication theoretical accounts this differentiation is ignored ( e.g. , “ A ” and “ a ” are considered the same ) . A more expansive theoretical account can easy suit this differentiation merely by handling capital and small letter characters as distinguishable symbols. Yet, from the input position, both attacks are incorrect. Uppercase and small letter characters are ne’er entered via separate keys on a keyboard ; therefore, the apparently more accurate intervention of capital and small letter characters as distinguishable symbols is merely as incorrect.

Dictionary theoretical account, a database of words with their frequence of usage, can besides be used for prognostic text input. In this instance, anticipation is done by seeking the lexicon for lucifers and telling by comparative frequence of words. Predictions made utilizing dictionary theoretical accounts are normally non every bit accurate as anticipations made from a principal. But the job with dictionary theoretical account is that it is memory intensive. As memory is both limited and expensive, dictionary theoretical account has a really important cost, both in footings of money paid for memory and what could hold been used to supply other characteristics like extra games, larger reference book, storage of more SMS messages, longer voice memos, screen rescuers, and so on, is lost in supplying linguistic communication dictionary ( R.Shriram et Al ( 2006 ) . However, sing the limited processing power of nomadic phones, and the fact that a principal might non reflect a user ‘s linguistic communication the dictionary theoretical account is largely used. Examples of dictionary-based Predictive Text solutions merchandises include T9 and eZiText.

Methodology

This paper implements text entry in three Nigerian linguistic communications for nomadic communicating utilizing prognostic multi-tap text entry system. The system is an application for making and directing text messages via SMS on nomadic phones. The system is aimed at cut downing the clip spent and besides, to ease the jobs encountered by Nigerian users in composing text messages in their native linguistic communication. The user composes text messages utilizing a prognostic text input system which is complimentary to the basic multi-tap text entry method most users are familiar with. It uses the text entry behavior of multi-tap while at the same clip supplying anticipation of whole words for the user to choose. For illustration, to compose the word “ Nigeria ” , prognostic multi-tap text entry system predicts the word for choice. The anticipation is done by looking up a lexicon of words for any lucifer with the substring. Wordss that lucifer are presented in a list ordered based on their chance of happening by frequence and recentness of usage, in combination with some conditions.

P ( word ) = Pfrequency ( word ) ten Precency ( word ) ( 2 )

Where P ( word ) , Pfrequency ( word ) and Precency ( word ) are chance of the word occurring, chance of the word utilizing comparative frequence and chance of the word utilizing recentness conditions, which are:

Wordss that start with the hunt substring are placed at the top of the list and are ordered by most likely to least likely. Most likely is given by argMax P ( words )

Wordss that do n’t get down with the hunt substring are placed at the underside of the list and are ordered by most likely to least likely.

If there are words in the same category ( words that begin with the hunt substring or words that do n’t get down with the hunt substring ) have equal chances. Then these words are ordered by length, with the greatest length preceding.

If no lucifer was found or the intended word is non in the anticipation list, the user can add the intended word to the system, therefore, bettering the public presentation of the system in future.

After the message has been composed, it is so sent via SMS to an reference ( phone figure ) specified by the user.

Consequences and Discussion

The text input theoretical account proposed above was simulated utilizing the Java Micro Edition. The consequences are shown as screen shootings.

When the application is launched, the compose message interface will be activated first, Figure 3 shows the message composing interface. This interface allows the user to type in the message utilizing multitap method. It besides has a anticipation bid which is used to put a demand on the system to foretell the word being typed. When the anticipation key is pressed, the predicted words list is displayed leting the user to choose the intended word. After choosing the intended word, the system automatically completes the word being typed. The composed message is sent by typing a valid phone figure in the finish reference interface and so pressing the Ok button.

.

Figure 3: Message composing interface

The user interface was designed to be user friendly, really simple and consecutive forward. Commands for operations explicitly depict what is being done.

Figure 4: Prediction list

In Figure 4, the user enters text utilizing multitap input method and presses the anticipation button, so that the system can foretell the word being typed. The user selects the coveted word or adds the intended word to the dictionary by pressing the attention deficit disorder word bid.

Figure 5: Menu options

From the compose message interface shown in Figure 5, the bill of fare contains five options. The continue option allows user to go on with the composing of their message. The Dictionary hierarchical menu is use to choose which of the three Nigerian linguistic communications to utilize. The “ My words ” hierarchical menu options allows users to add new words, to cancel a word, and to redact a word. The Help submenu provides users with contextual aid, while Exit allows users to go out from the bill of fare.

Figure 6: Dictionary choice

From the dictionary list presented as shown in Figure 6, the coveted lexicon is selected and the OK button pressed.

Decision

From this paper, it can be seen that uniting multitap input technique with prognostic text input can assist cut down the figure of key strokes per missive, thereby increasing the rate at which text is entered, and cut downing emphasis associated text input on nomadic devices. In Nigeria, applications developed on this paper will assist increase the day-to-day usage of our female parent lingua and can assist other people who are non fluid with the linguistic communications ( Hausa, Igbo and Yoruba ) communicate better by cut downing their spelling errors.

To develop applications for usage in Nigeria, the followerss are needed to be included in the input technique, anticipation of when last the word was used, anticipation of the following word to be typed, and phrase anticipation. Besides, awaited addition in the processing power of future nomadic devices by utilizing a principal based prognostic theoretical account which will give better anticipations can be implemented.