Title of Invention

TEXT INPUT DEVICE FOR INDIC LANGUAGES

Abstract

We claim: 1.A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts comprising input keys that are blocked into consonant block, vowel block, semi¬vowel block and others such as punctuation, function keys; wherein keys in the consonant and semi-vowel blocks are organized in alphabetic order based on their phonetic classification wherein keys as Matras (vowel modifiers) arranged in the vowel block, as per their normal graphical location in a character with keys for other modifier graphemes that are perceived by the users to be similar to vowel matras (&, $P) located within the vowel block wherein other keys such as punctuations, function keys and their like being located in clusters and retained as in traditional QWERTY layout.

Full Text

FORM - 2 THE PATENTS ACT, 1970 (39 OF 1970) COMPLETE SPECIFICATION (See section 10; rule 13) TITLE OF INVENTION "Text Input Device for Indie Languages" (a) INDIAN INSTITUTE OF TECHNOLOGY Bombay (b) having administrative office at Powai, Mumbai 400076, State of Maharashtra, India and (c) an autonomous educational Institute, and established in India under the Institutes of Technology Act 1961. The following specification particularly describes the nature of the invention and the manner in which it is to be performed. ORIGINAL GRANTED 15-7-2007 Field of the invention This invention relates generally to a new and effective data entry device such as a keyboard for inputting'Indie text scripts such as Devnagari text script and the similar phonetic scripts. Background of the invention The need for simple and easy to use text input devices for Indie languages is an imperative as the IT revolution is sweeping India including its towns and villages. Surprisingly most computers in India still have primary input mechanism in English. Use of computers in India can be enhanced if computing devices are made available with easy input of Indie text scripts. Keyboards were originalfy designed to input text in the Roman script. The Roman script can be almost completely represented by 26 keys on the keyboard. Each lower case character is achieved by one keystroke. Each upper case character is available on the 'shifted' position of the corresponding lower case key. This arrangement poses almost no cognitive load on even the first time users. Unfortunately, this ease of use is not applicable to Indie scripts. Three issues related to Indie scripts pose challenges for the design of input devices: • Structure of Indie scripts • The difference between the cognitive styles of writing and typing • Large number of characters Structure of Indie Scripts The structure of Indie scripts, for example Devnagari, is such that users need to type multiple keystrokes on a keyboard to enter one character in the computer. Barring a few exceptions, the structure is common for all Indie scripts, though the scripts themselves may be distinctly different. Devnagari has 53 basic letters - 34 consonants and 19 vowels in addition to numbers and punctuation marks. The last nine of the 34 consonants are also known as semi-vowels. One or more letters come together to form a character and one or more such characters come together to form a word. Letters can be combined in different ways to form characters. It would be relevant to survey some of the possible variations and combinations in Devnagari scripts: • (C+V): A simple character consists of a combination of a consonant and a vowel (C+V). For example, the word tTFfr (paanee) is made up of two characters trr (pa) and sft (nee). The character m (pa) is made up of the consonant *T (p) and the vowel 3TT (aa), and the character ^t (nee) is made up of the consonant £T (n) and the vowel f (ee). msft = m + sfr = q" + arr + ^r + f paanee = paa + nee = p + aa + n + ee (C+C+V): A more complex character (called conjunct) would consist of two or more consonants combined with a vowel (C+C+V). For example, the character ft (pre) in the word ft^r (prem) is made up of conjunct *T (p), the conjunct T (r) and the vowel 7 (e). ft^T = ft + 3T = V + J + T? + 3T + 3T prem = pre + m = p + r + e + m + a • (C+V+V): In some cases, two vowels are combined with one or more consonants (C+V+V), For example in the word trfe (paande), the character W (paan) is made up of the consonant IT (p), the vowel 3TT (aa) and the vowel 3T (an). tri£ = HT + £ = TT + 3TT + 3T+ ? + ^ paande = paan + de = p + aa + an + d + e Cognitive Styles of Writing: and Typing In a few characters, there are differences in the phonetic (or spoken) sequence of consonants and vowels and the visual (or written) sequence of their corresponding glyphs. • For example for the character fft (pi) in the word fftcTT (pitaa) some users write the glyph (f^) for the vowel ^ (i) before the glyph for the consonant q; (p), the f (i) is always pronounced after the *T (p). fft^TT = fa + ?TT = H + ? + H + 3TT pitaa = pi + taa = p + i + t + aa • In another example, in the word 3T25" (arth), users write the glyph for the consonant 21 (th) before the glyph for the consonant T (r) though the sequence of pronunciation of the character is sfr (rth). 3ftf = 3T + 2t = 3T + T + 2T + 3T arth = a + rth = a + r + th + a Some users are habituated to thinking in the visual sequence of the glyphs as they type. Other users are habituated to thinking inispoken sequences as they type. Though occasional, these differences are significant enough to baffle novices, require learning and put an extra cognitive load on users during typing. Theoretically, infinite number of Devnagari characters can be generated by combinations of C+V, C+C+V C+C+...V... structures. Actually, a smaller subset of characters is currently used in Devnagari. But even this small subset is quite1 large making it impossible^ design a practical device such as a keyboard in which one key corresponds to only one character. It is necessary to have multiple keystrokes of the base consonants and vowels to generate most Devnagari characters. However, sufficient number of keys are not available on QWERTY keyboards even to represent the basic letters - 34 consonants, 19 vowels and the halant key which require 54 keys against the 26 that are available. All software solutions currently available overcome this problem by using shifted positions of keys for some letters. Hence, typing in Indie scripts presents a significant cognitive load on the users. Current solutions do not adequately address this problem. As a result, long training hours are currently needed to learn to type in Indie scripts. It is estimated that it takes 30-50 hours of training and practice to achieve typing speeds of 25 words per minute (WPM) or about 125 characters per minute (CPM). This discourages most people to learn to type in Indie scripts. Most people who type in Indie scripts are professional typists who have undergone formal training. The present invention overcomes problems of the present day data input devices for scripts such as Indie Scripts and provides keyboards design for easy input of Indie text with minimal cognitive load on the learner. Prior art Several software solutions based on QWERTY have been proposed or are in practice for the input of Indie languages. These are some of the options: 1. Inscript (e.g. Leap Office, C-DAC), The Bureau of Indian Standards adopted ISCII and developed a documentation IS13194: 1991 2. English phonetic (e.g. Modular Systems, http://www.aksharamala.com/help/chm/lnDut%20Schemes/ITRANS/Devan aqari/kev combo.html 3. Layout in commercially available typewriter (e.g. Godrej and Remington; http://www.devyani.com/dl/godrei.pdf 4. ITR (by Indian Typographic research) 5. Desha ("Desha keyboard" Multilingual Phonetic Keyboard designed and developed by National Centre for Software Technology, Mumbai, India, NCST Report by Prof R K Joshi, 9th February, 2001) 6. Transliteration (e.g. BIS std.: followed by portals like e-patra.com, webduniya.com etc.) The main limitation of devices and approaches described above is that the keyboards used are not explicitly designed for the input in Indian languages. Inscript keyboard layout [Figure 11 The Inscript overlay can be used on any QWERTY keyboard. Activating the Caps-Lock selects the Inscript overlay; otherwise, normal lower case English overlay gets selected. The upper image (figure 1) shows the layout of keys in the normal (un-shifted) mode, whereas the lower image (figure 1) shows the layout of keys in the shifted mode. The INSCRIPT keyboard overlay was standardized by Ministry of Information Technology, Government of India 1986. A Department of Electronics committee did a revision in 1988. Currently this is the most popular keyboard layout for Indie languages. Desk Top Publishing (DTP) operators use it as most of the word processing packages offer this as the default layout. Following are the weaknesses of the Inscript layout: • New users face difficulty in locating keys in the shifted position. E.g. oT oT (normal) and ^T (shifted) share the same key, and new users take more time to find ^T than 3T. • There are inconsistencies in the grouping of keys resulting in extended learning times. Keys for four consonants in a group of five are together, while key for one consonant is away. E.g. keys for ^, m, ~$ and 3T are together, but the key for the letter 3T is away. Similarly is the situation for other groups of consonants. • There are inconsistencies in the typing sequences that lead to learning difficulties. E.g. cjT + at give cfit. But 3T + at give 3Tat and not 3ft as expected. To type 3ft", one has to press a special key. • 2J", though a frequently used key, is in an unlikely position (question mark). • Inscript favours touch typing, which is for trained professionals. However but most non-professional users type with a few fingers. Phonetic layout Tfiaure 21 This layout is based on the phonetic equivalents of Roman letters on the QWERTY keyboard. Typically, the letter q" is located on the same position as the letter p, whereas the letter ^ is located on capital P. Following are the drawbacks in the Phonetic layout: • The keyboard relies on the knowledge of typing in the Roman script. Only 11 % of Indian population is familiar with the English language and the Roman script. Fewer still are familiar with typing in the Roman script. • In Devnagari, there are multiple letters corresponding to a letter in the Roman script. For example, the phonetic equivalent of the letter t could be cT and Z. Similarly, the phonetic equivalent of the letter d could be Z. and 5. This problem has been solved in this layout by making some arbitrary mappings. So Z is mapped on ' (single quote) and 5 is mapped on [ (square open bracket). This arbitrariness adds to learning difficulties, even for those who are familiar with typing in Roman script. • Some other mapping that is arbitrary includes frequent letters such as 3TT (F), ? (z), f (Z), 3 (q), 2F (Q), * (w), * (W), 3TT f), 3ft (~). These arbitrary mappings effectively seem to nullify any learnability advantage that may be otherwise afforded on people familiar with Roman typing. Typewriter layouts Ffiqures 3-41 These layouts are similar to the Hindi typewriter layout and useful for typists familiar with Typewriter layout. The Keyboard Layout and the Key Sequence Charts shown in the figure 3 and 4 can be used to find the correct key combinations. Following are the drawbacks in these layouts: • Keys are not grouped in any particular order readily discernable to users leading to long response time in locating keys. • The keyboards require users to differentiate between the antyadand-the vertical line that comes in the end of consonants like ST - and the matra or in the characters like £TT. This is confusing to a beginner. • Typing in these layouts is not compatible with Unicode, ISCII and other standard storage formats. Indian Typographic Research (ITR) layout ffiqure 51 The Keyboard Layout shown in the figure 5 can be used to find the correct key combinations. Currently this is not widely used as it requires intense training. Following are the drawbacks found in the ITR layout: • Keys are not grouped in any particular order readily discernable to users, leading to long response time in locating keys. • • To type f one needs to type ^ followed by a rafar character ( ) which is not consistent with the users' understanding of the script. • Typing in ITR layout is not compatible with Unicode, ISCII and other standard storage formats. Chinese Patent CN1341885 describes a two-key input method to input character by using numeric key-press. The numbers, all English letters and characters on standard keyboard or Chinese phonetic letters can be inputted by the method using numeric key-press. It is best suited for Chinese and Korean scripts, which are ideographic, and each of the character is semanticaily distinct and complex. The solution offered in this patent cannot be adapted to Indie scripts. US Patent US6009444 discloses a text input device (10) and a method of data entry for entering text in a language (e.g. Zhuyin phonetic Chinese), in which a character is represented as a first symbol selected from a first subset of symbols and a second symbol selected from a second subset of symbols, where the first and second subsets are mutually exclusive. A first key on which is displayed a first subset of symbols is activated (e.g. any one of keys 1-6). A candidate first symbol is displayed in response to the step of activating the first key. A second key is activated on which is displayed a second subset of symbols (e.g. any one of keys 7-0). The candidate first symbol is fixed and a candidate second symbol is displayed in response to activating the second key. A third key can be activated (e.g. any one of keys 7-0), on which is displayed a further subset of symbols, whereupon the candidate second symbol is fixed. This invention suggests the subsets and activity of searching with gradual narrowing of focus, but has no relationship to Indie scripts or their structure. Moreover such an approach does not ease the cognitive load on the learner. The prior art described above is replete with examples that solve the problems of limited keyboard area, but in turn overburden the learner with substantial cognitive load leading to slow learning with the risk of committing mistakes while text inputting based on Indie Scripts. Detailed description of the invention The invention is now described and illustrated with non-limiting examples. Brief description of the drawings Figure 1: Shows the Inscript keyboard layout Figure 2: Shows the English phonetic keyboard layout Figure 3: Shows the Remington Typewriter Layout adapted for a computer keyboard Figure 4: Shows the Godrej Typewriter Layout adapted for the computer keyboard Figure 5: Shows the Indian Typographic Research (ITR) layout Figure 6: Shows the front view of one embodiment of invention, called Barakhadi 1, and its keyboard layout. Figure 7: Illustrates the "Varnamala" showing the alphabetic order and the structure of Devnagari script in terms of vowels and consonants. Figure 8: Shows a closeup of the consonant block of the Barakhadi 1 keyboard. Figure 9: Shows a closeup of the vowel block of the Barakhadi 1, with matra (vowel modifier) keys in a character. Figure 10: Shows Barakhadi 2, another version of the embodiment of the invention, with 5 rows of keys to make suitable for desktop application with manageable span of keys. Figure 11: Shows the comprehensive view of Barakhadi 2. Figure 12: Shows a closeup of the vowel block in Barakhadi 2. Figure 13: Shows an embodiment in which the layout in which the positions of consonant and vowel blocks are exchanged. Figure 14: Shows on-screen version of the keyboard for the kiosk based interaction where the typing is performed on touch screens. • Figure 15: Shows an embodiment of the invention, called Keylekh 3, which also supports the typing the Roman script through QWERTY layout mapped on the lower three rows to facilitate bi-lingual typing. • Figure 16: Shows an embodiment of the invention in which the function keys and punctuation keys are shifted on top to achieve the standard length of the keyboard. • Figure 17: Shows an embodiment of the invention, where the consonant, vowel and semi-vowel blocks are mapped on the standard QWERTY keyboard with the use of upper Number keys row and Function keys row. • Figure 18: Shows results of tests comparing learnability of Barakhadi keyboard with Inscript keyboard by two Hindi school teachers. • Figure 19: Shows results of tests conducted on six learners with basic literacy levels. A comparison is made of the learnability using Keylekh keyboard with that of Inscript keyboard. • Figure 20: Shows results of tests conducted on eight learners with advanced technical education to compare the learnability of Keylekh keyboard with Inscript keyboard. In accordance with this invention the keyboard layout is as follows: • Keyboard consists of these blocks of keys: consonant block, vowel block, semi-vowel block and others such as punctuation, function keys etc. • The keys in the consonant and semi-vowel blocks, are organized in the easy to find alphabetic order, which has its basis in their phonetic classification. • Matras (vowel modifiers) are arranged in the vowel block, as per their normal graphical location in a character. Further the, keys for some other modifier graphemes that are perceived by the users to be similar to vowel matras (&, W) are also similarly located within the vowel block. • The other keys (e.g. punctuations, function keys etc.) are in clusters and retained close to traditional QWERTY layout. In one of the embodiments of this invention, Barakhadi 1, as illustrated in figure 6, 8, and 9: • The left-most block is the consonant block, organized in the easy to find alphabetic order (figure 6). • The central block is the vowel block (figure 9). In the vowel block, matras (vowel modifiers) are arranged as per their normal graphical location in a character. In addition, some keys for graphemes that are perceived to be like matras (&, sP) are also similarly located. In another embodiment of this invention, Barakhadi 2, as illustrated in figure 10, 11 and 12, the essential nature of consonant, vowel and semi-vowel blocks described above is maintained, but the blocks themselves are rearranged to ensure that the number of rows on the keyboard is restricted to five. • The consonant block is kept to the left hand side while semivowel block is shifted to make it fit in five rows (figure 10). • The semi-vowel block is moved to the top-central position, and the vowel block is moved down, (figure 12) • The keys in the vowel block are painted with a different colour and alignment is altered to differentiate this block from the semi-vowel block. • In the vowel block, matras are arranged as per their normal graphical orientation in a character with respect the 'central' key for the vowel 3T. Here the 3T key acts as a reference, and hence is painted differently. In another embodiment of this invention, as illustrated in figure 13, the positions of consonant block and semi-vowel + vowel blocks are exchanged. This layout brings in the more frequently used keys to the centre of the keyboard and pushes the relatively less-frequently used keys to the sides. In addition, this version supports typing the Roman script. In another embodiment of the invention, as illustrated in the Fig 14, on-screen version of the keyboard for the kiosk based interaction where the typing is performed on touch screens. Another embodiment of the invention, Keylekh 3, as illustrated in the Fig 15, supports quick typing of both Roman and Devnagari scripts. In this version, the keys in the consonant, semi-vowel and vowel blocks are staggered to match standard QWERTY keyboards. This facilitates quick typing in Roman script by users who are already familiar with the QWERTY keyboards. In another embodiment of the invention, as illustrated in the Fig 16, the function keys and the punctuation keys are located above the consonant, semi-vowel and vowel blocks. This ensures that the width of the keyboard is the same as standard QWERTY keyboards. In another embodiment of the invention, as illustrated in the Fig 17, the consonant, semi-vowel and vowel blocks layout is mapped on the QWERTY keyboard with the use of upper Number keys row and Function keys row and with the help of colour coding of keys. This facilitates the English typing through the regular three rows. All Indie scripts have similar structure, with some additional vowels and / or consonants. Though this invention has been demonstrated on Devnagari, the concept can be easily extended to other Indie scripts. The invention can run in multiple 'modes' to be compatible with software applications capable of accepting different scripts and different keyboard layouts as inputs. For example, the keyboard Keylekh 3 can currently run in two modes -Devnagari Inscript mode and QWERTY Roman mode. In Devnagari Inscript mode, it is compatible with software applications that accept Devnagari Inscript input. For example, in the Devnagari Inscript mode, the z key sends out the code for' (single quote), which Devnagari Inscript interpreters read as Z. In QWERTY Roman mode, it is compatible with all applications accepting QWERTY Roman input (like most standard keyboards). In the same example above, in QWERTY Roman mode, the same Z key sends out the code for lower case u, because Z key is in approximately the same physical location where lower case u would fall in standard QWERTY keyboards. This is realized by building a special controller circuit in the keyboard with a microprocessor and memory for a look-up table for all keystrokes in each mode. After each keystroke, the data is pre-processed by this microprocessor before it is sent to the computer so that the data being sent out is compatible with existing mode on the computer. This system can be extended to support a wide range of input scripts and layouts such as QWERTY, Inscript, Desha, Phonetic, ITR etc. The keyboards of the present invention were tested on diverse subjects to establish their applicability. Experiments were conducted to compare the effectiveness of Barakhadi 1 (figure 6) with standard Inscript keyboards. Two school teachers proficient in Hindi but novices in typing were given one keyboard each and their typing speeds were monitored over a set of 8 tasks. The results show that Barakhadi 1 did significantly better than Inscript especially during the initial learning stage of typing (figure 18). Experiments were conducted to compare the effectiveness of a variant of Keylekh 3 (figure 18) with standard Inscript keyboards. Three persons with basic literacy levels but novices in typing were given Keylekh keyboard. Three other persons at similar literacy level but novices in typing were given Inscript keyboard. Their typing speeds were monitored over a set of 18 tasks. The results showed that Keylekh keyboard did consistently better than Inscript, particularly in the initial stages (figure 19). Experiments, were conducted to compare the effectiveness of a variant of Keylekh 3 keyboard (figure 18) with standard Inscript keyboards. Four students with advanced technology education but novice in Hindi typing were given Keylekh keyboard and four students with similar background were given Inscript keyboard. Their typing speeds were monitored over a set of 18 tasks. The results showed that Keylekh keyboard did consistently better than Inscript, particularly in the initial stages (figure 20). These unequivocally establish that the keyboards of the present invention are effective devices for easy inputting of Devnagari text scripts especially for first learners of keyboard users. The advantage of the arrangement of keys into consonant, semi-vowel and vowel blocks in the manner described above is that the users can locate the keys easily in familiar, predictable locations. This reduces the cognitive load on users' mind as he learns to input Devnagri text using this device. The Varnamala-based layout provides a cognitive reference for the key search and thus makes it intuitive and easy-to-use for learning to type in Indie scripts. Keylekh can have direct application to all languages that use Devanagari script such as Hindi, Marathi, Rajasthani, Sanskrit and Konkani. With minor modifications, it is also easily applicable to Indie scripts derived from the Bramhi script such as Gujarati, Punjabi, Bengali, Assamese and Oriya. With minor modifications, it could be extended to accommodate other scripts related to the Bramhi script such as Kannada, Telugu, Tamil, Malayalam, Thai, Kashmiri, Persian and Urdu etc. The concept of using a familiar grouping of letters in the alphabet as a layout on the keyboard can be extended to any script that has such a grouping of letters in the alphabet that users of that script can identify. Claims We claim: 1. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts comprising input keys that are blocked into consonant block, vowel block, semi¬vowel block and others such as punctuation, function keys; wherein keys in the consonant and semi-vowel blocks are organized in alphabetic order based on their phonetic classification wherein keys as Matras (vowel modifiers) arranged in the vowel block, as per their normal graphical location in a character with keys for other modifier graphemes that are perceived by the users to be similar to vowel matras (&, $P) located within the vowel block wherein other keys such as punctuations, function keys and their like being located in clusters and retained as in traditional QWERTY layout. 2. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein: • the left-most block is a consonant and semi-vowel block, organized in the easy to find alphabetic order • the central block is a vowel block in which the matras (vowel modifiers) are arranged as per their normal graphical location in a character and keys for graphemes that are perceived to be like matras {%>, ?T>) are similarly located. 3. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the blocks are rearranged so that the number of rows on the keyboard is restricted to five with the consonant block located on the left side while semivowel block is shifted to make it fit in five rows. 4. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the semi-vowel block is moved to the top-central position, and the vowel block is moved down. 5. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the keys in the vowel block are painted with a different colour and alignment is altered to differentiate this block from the semi-vowel block. 6. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein, within the vowel block matras are arranged as per their normal graphical orientation in a character with respect the 'central' key for the vowel 3T. Here the 3J key acts as a reference, and hence is painted differently. 7. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 and 4 wherein the positions of consonant block and semi-vowel + vowel blocks are exchanged so that the frequently used keys are brought to the centre of the keyboard and relatively less-frequently used keys are located on the sides. 8. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein on-screen version of the keyboard for the kiosk based interaction where the typing is performed on touch screens 9. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the keys in the consonant, semi-vowel and vowel blocks are staggered to match standard QWERTY keyboards as illustrated in the thereby simultaneously supporting quick typing of both Roman and Devnagari scripts. 10. A data entry device such as a keyboard for inputting Indic text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the function keys and the punctuation keys are located above the consonant, semi-vowel and vowel blocks. 11. A data entry device such as a keyboard for inputting Indie text scripts such as Devanagari text script and the similar phonetic scripts as claimed in claim 1 wherein the consonant, semi-vowel and vowel blocks layout is mapped on the QWERTY keyboard with the use of upper Number keys row and Function keys row and with the help of colour coding of keys. Dr.Prabuddha Ganguli Agent on behalf of Applicant

Documents:

720-mum-2003-cancelled pages (15-06-2007).pdf

720-mum-2003-claim(granted)-(15-6-2007).doc

720-mum-2003-claims(granted)-(15-06-2007).pdf

720-mum-2003-correspondence (15-06-2007).pdf

720-mum-2003-correspondence(ipo)-(07-01-2008).pdf

720-mum-2003-form 1(15-07-2004).pdf

720-mum-2003-form 19(15-07-2004).pdf

720-mum-2003-form 2(granted)-(15-06-2007).pdf

720-mum-2003-form 2(granted)-(15-6-2007).doc

720-mum-2003-form 26(15-07-2004).pdf

720-mum-2003-form 3(15-07-2004).pdf

720-mum-2003-form 3(16-07-2003).pdf

720-mum-2003-form 5(15-07-2004).pdf