1. Thesis Presented By
Mohammad Abul Kalam Azad C Shabbir Ahmad C Francis Palma Tony C Supervised by S. M. Kamruzzaman Assistant Professor
Department of Computer Science and Engineering International Islamic University Chittagong

2. An Efficient Technique for Text Compression

3. Summery of Presentation
☼ Introduction. ☼ Methodology. ☼ How Many Words. ☼ Word Lookup Table. ☼ Word Storing Architecture. ☼ Memory Allocation Method. ☼ Memory Space Requirement. ☼ Data Management Algorithm. ☼ Data Compression. ☼ Experimental Result. ☼ Conclusions.

4. Introduction ☼ Data Storage. ☼ Information Management Using Symbol.
☼ Secret Data Communication.
☼ Storage Requirement Reduction Problem.
☼ Reduction.
☼ Compression.

5. Methodology The data compression will be done in two phases:
☼ Reduction Using Lookup Table.
☼ Compression Using Deflate Algorithm.

6. How Many Words in English
Webster’s Third New International Dictionary
470,000 entries.
The Oxford English Dictionary, second Edition, reports that it’s include similar number.
Including: Abbreviation, Phrase, Taboo words, Dialects, Family words.
Excluding: Names of entirely scientific terms.

7. Word Lookup Table
A word lookup table is a special tabular data file containing the text dimension of a word as an attribute of an address, which is used to pop up text to display the possible text data.

8. Word Lookup Table
A lookup table is defined by its capability of addressing. Expressed in bits.
2^ 19 =
That is we need 19-bit lookup table.

9. Figure 1. The architecture of the stored word in lookup table
Word Storing Architecture
Index Address
Stored Word in Details
Ending Signal
19-bit Address
7 Character long word
Termination signal
19 bit
7*7 = 49 bit
7 bit
Total 75 bit
Figure 1. The architecture of the stored word in lookup table

10. Word Storing Architecture
Table 1. Words in word lookup table
Index Address
Word stored in details
Ending signal
19-bit address a … ..
and
Computer
...
Intelligent

11. Word Storing Architecture
Table 2. Special situation handling addresses
Index address
Word stored in details
Ending signal
01--10
Termination of address
01--01
Single Uppercase
Multiple Uppercase
01--11
Multiple Uppercase Ter.
01--00
Title Case
Single tOGGLE cASE
Multiple tOGGLE cASE
Multiple tOG. cAS. Ter.

12. Word Storing Architecture
Table 3. Entry of different punctuation signs
Index address
Word stored in details
Ending Signal …
01--00 .
01--01
.+Backspace
11--10 ,
11--11
,+Backspace

13. Example Example 1 Example 2 Gen. He is a very good boy . 176 LUT 19
133
Example 2
Gen.
Sometimes I
need
some
helping
too .
248
LUT 19
133

14. Example Example 3 Example 4 Gen. Although computers may have basic
similarities ,
376
LUT 19
133
Example 4
Gen.
Several
systematic
tabular
methods
For
machine
reduction
exits .
512
LUT 19
171

15. Memory Space Requirement
Space = 2^19 * 75 bits
= * 75 bits
= bits
= Bytes
= 4800 Kilo Bytes
= Mega Bytes

16. Memory Allocation Method
Table 4. The Hash table
Index address
Starting Character a
and …
.....
buy
...
yolk
zoo

17. Figure 2: How the word lookup table will be stored
Memory Allocation Method
Figure 2: How the word lookup table will be stored
11--01
and
11--10
Andiron ……
00--01
sun
00--10
Sun-bath

18. Data Management Algorithm
Algorithm UnRedToRed( )
1. Read file
2. Read character to form a word until empty.
3. Finds its appropriate address from Hash table.
4. Find the word in Lookup Table.
5. If found then
6. Check case
7. If case = lower then
8. Fetch addresses
9. else
Do the case management
Fetch Address
12. Print the address
13. else
14. Give termination symbol.
15. Start ASCII storage (word)
16. Go to step 1.
End.

19. Data Management Algorithm
Algorithm RedToUnRed ( )
1. Read file
2. Fetch address.
3. Check Address status.
4. If word then,
5. Print the word.
6. If situation handles then,
7. Do according to it.
8. Go to step 2.
End

20. Data Compression Methods
☼ Lossy Data Compression.
☼ Lossless Data Compression.

21. Lossless Data Compression
☼ Run Length Encoding (RLE).
☼ Huffman Coding.
☼ Lempel-Ziv 77 Encoding (LZ77).
☼ Deflate Algorithm.

22. Run Length Encoding <Esc> Specific Character <Frequency>
XXXXXXXXXXXXXXX that’s all, folks!
<Esc> X <15> that’s all, folks!

23. Huffman Coding Finally we get a new binary code for each character:
Let a sequence of character with their frequencies are:
A B C D E F G
Finally we get a new binary code for each character:
F : B : 01
C : A : 101
G : D : 1110
E : 1111

24. Lempel-Ziv 77 Encoding
the_rain_in_Spain_falls_mainly_in_the_plain
the_rain_
the_rain_<3,3>
So, in binary, the pointer <3,3> would look like this:
the_rain_<3,3>Sp
the_rain_<3,3>Sp<9,4>
the_rain_<3,3>Sp<9,4>falls_m<11,3>

25. Combination of Huffman coding and Lempel-Ziv 77 encoding
Deflate Algorithm
Combination of Huffman coding and Lempel-Ziv 77 encoding

26. Experimental Result

27. Comparison with other methods
☼ In general compression rate from 12% to highest 50%.
☼ Proposed method 53% reduction

28. According to Proposed Method + Deflate Algorithm
Comparison with other zip software
Compression Type
Size
Normal
78.53 KB
According to Proposed Method + Deflate Algorithm
14.38 KB
Gzip
29.61 KB
Winzip
31.27 KB

29. Comparison with other zip software

30. Conclusion ☼ Text Data Storage Reduced to 53%.
☼ After Compression 75% - 80%.
☼ Faster Portable.

31. Questions

32. Thank You