User:Krvoje/Application2012 - Revision history

Krvoje: /* Bio */

2012-04-05T19:10:51Z

Bio

Krvoje at 18:34, 5 April 2012

2012-04-05T18:34:07Z

Krvoje: /* The coding period: */

2012-04-05T15:08:41Z

The coding period:

Krvoje: /* The coding period: */

2012-04-05T14:53:28Z

The coding period:

Krvoje: /* The coding period: */

2012-04-05T14:28:18Z

The coding period:

Krvoje: /* The coding period: */

2012-04-05T14:27:48Z

The coding period:

Krvoje: /* The coding period: */

2012-04-05T14:02:35Z

The coding period:

Krvoje: /* Bio */

2012-04-05T14:01:32Z

Bio

Krvoje: /* What do you plan to do? */

2012-04-05T14:00:12Z

What do you plan to do?

Krvoje: /* The coding period: */

2012-04-05T13:59:07Z

The coding period:

@@ Line 210: / Line 210: @@
 Regarding the technologies used in machine translation I've been enrolled in courses with finite state machines, and context free grammars (implementation of a parser using yacc+flex), and machine learning.
-==References==
+=References=
 <references/>

@@ Line 1: / Line 1: @@
-(This is a draft)
 == GSoC application: Rule-based finite-state disambiguation ==
 Hrvoje Peradin

@@ Line 177: / Line 177: @@
 ==The coding period:==
-- Week 1-2: Writing prototypes of rules in foma, test different approaches and minimization alogrithms to see which give the best performance
+- Week 1-2: Writing prototypes of rules in foma, test different approaches and minimization algorithms to see which give the best performance
 - Week 3-4: Writing and testing additions to lttoolbox for compiling FST's from finite-state calculus, based on the results from the testing
@@ Line 189: / Line 189: @@
 - Week 6-7: Writing the compiler for the XML format with the new modules of lttoolbox API, and a stream processor for the compiled rules
-- Week 8-9: Writing rule examples and testing them with the compiler and the stream processor
+- Week 8-9: Writing rule examples and testing the compiler and the stream processor with them
 - Week 10: Writing use-case examples, based on the final format of the XML formalism

@@ Line 179: / Line 179: @@
 - Week 1-2: Writing prototypes of rules in foma, test different approaches and minimization alogrithms to see which give the best performance
-- Weeks 3-4: Writing and testing additions to lttoolbox for compiling FST's from finite-state calculus, based on the results from the testing
+- Week 3-4: Writing and testing additions to lttoolbox for compiling FST's from finite-state calculus, based on the results from the testing
 - '''Deliverable #1''': The module for lttoolbox for parsing a finite-state calculus (compatible with foma and xfst) and compiling FST's.
-- Week 5-6: Writing the description of the XML formalism and writing a validator for it, expressing the rules in finite-state calculus.
+- Week 5: Writing the description of the XML formalism and writing a validator for it, expressing the rules in finite-state calculus.
 - '''Deliverable #2''': A complete XML formalism for expressing finite-state disambiguation rules, with validation and formulation of rules in finite-state calculus.
-- Week 7-8: Writing the compiler for the XML format, and a stream processor, testing the software.
+- Week 6-7: Writing the compiler for the XML format with the new modules of lttoolbox API, and a stream processor for the compiled rules
+- Week 8-9: Writing rule examples and testing them with the compiler and the stream processor
 - Week 10: Writing use-case examples, based on the final format of the XML formalism

@@ Line 181: / Line 181: @@
 - Weeks 3-4: Writing and testing additions to lttoolbox for compiling FST's from finite-state calculus, based on the results from the testing
-- '''Deliverable #1''' : The module for lttoolbox for parsing a finite-state calculus (compatible with foma and xfst) and compiling FST's.
+- '''Deliverable #1''': The module for lttoolbox for parsing a finite-state calculus (compatible with foma and xfst) and compiling FST's.
 - Week 5-6: Writing the description of the XML formalism and writing a validator for it, expressing the rules in finite-state calculus.
-- '''Deliverable #2''' : A complete XML formalism for expressing finite-state disambiguation rules, with validation and formulation of rules in finite-state calculus.
+- '''Deliverable #2''': A complete XML formalism for expressing finite-state disambiguation rules, with validation and formulation of rules in finite-state calculus.
 - Week 7-8: Writing the compiler for the XML format, and a stream processor, testing the software.
@@ Line 195: / Line 195: @@
 - Week 12: Final cleanup of the code, and the documentation
-- '''Deliverable #3''' : The complete disambiguation system, with a compiler, a processor, and the documentation containing use-case examples.
+- '''Deliverable #3''': The complete disambiguation system, with a compiler, a processor, and the documentation containing use-case examples.
 ==Non-GSoC activities==

@@ Line 187: / Line 187: @@
 - Week 7-12: Writing the compiler for the XML format, and a stream processor, testing the software.
-- '''Deliverable #3''' : The complete disambiguation system, with a compiler, a processor, and the documentation.
+- '''Deliverable #3''' : The complete disambiguation system, with a compiler, a processor, and the documentation containing use-case examples.
 ==Non-GSoC activities==

@@ Line 200: / Line 200: @@
 I have worked on the language pair apertium-sh-mk for the GSoC of 2011., and have been a mentor for Google Code-In 2011 for several tasks involving that and similar language pairs.
-Regarding the technologies used in machine translation we I've been enrolled in courses with finite state machines, and context free grammars (implementation of a parser using yacc+flex), and machine learning.
+Regarding the technologies used in machine translation I've been enrolled in courses with finite state machines, and context free grammars (implementation of a parser using yacc+flex), and machine learning.
 ==References==

@@ Line 36: / Line 36: @@
 Although I am already quite familiar with the modules of lttoolbox API for compiling and operating on transducers, and these are the parts that will be the focus of my work, the API is rather sparsely documented, and there is a factor of impredictability in how much time certain tasks will consume. For this reason I have assigned a larger time frame to the part of work regarding the API.
-The disambiguation module will be documented with use-case examples in various languages, based on my current experience with developing Constraint Grammar for Croatian.
+The disambiguation module will be documented with use-case examples of varying complexity in various languages, based on my current experience with disambiguating Croatian texts with Constraint Grammar.
 ===Examples of rule formalism===

@@ Line 178: / Line 178: @@
 - Weeks 1-4: Writing and testing additions to lttoolbox for compiling FST's from finite-state calculus
-Since expressing disambiguation rules requires more complicated regular expressions than are currently supported in lttoolbox, I will write a module for compiling a larger subset of finite-state calculus to lttoolbox transducers. The formalism will be compatible with foma/xfst.
-Lttoolbox is sparsely documented and it is difficult to predict the exact time I will spend writing the additions, so for this reason I am assigning such a large time frame to this part of the task.
 - '''Deliverable #1''' : The module for lttoolbox for parsing a finite-state calculus (compatible with foma and xfst) and compiling FST's.
@@ Line 191: / Line 188: @@
 - '''Deliverable #3''' : The complete disambiguation system, with a compiler, a processor, and the documentation.
-For testing of the rules I will use examples of rules of varying complexity based on my current experience with disambiguating texts.
-The documentation will contain use-case examples in form of tutorials, and will be based on my current work with Constraint Grammar for disambiguation of Croatian texts. The examples will be based mostly on the rules I am currently developing for Croatian, as well as on Constraint Grammar rules available for other languages.
 ==Non-GSoC activities==