December 11

Having finished up my slides very early that morning, I rested up for my final presentation. Overall it went well, however it wasn't perfect. I kinda got a bit flustered and forgot that I had made notes on what I wanted to say for each slide. Whoops..... Now all I have left to do is finish up my paper. It shouldn't' be too bad, building the presentation made me think about everything and so writing the paper should just be putting in down on paper instead of slides now.

December 2; 2:00AM

The start of a beautiful half birthday....
I have fully worked out the major kinks in my first parallel implementation of Earley's algorithm, and it is running with full success. Now all I have left to do is clean up the code a bit and run some actual tests to see if i gained anything at all with parallelization.

November 12

Have worked out all the kinks in my Earley parser, and it works successfully. now my task is to dive into MPI enough to code a parallel Earley algorithm. Additionally I need to develop a series of test cases that are taxing enough that they would warrant parallelization.

November 5

Worked on writing a serial Earley parser that has the table partitioned according to the start and end position of the rules. Got basic framework done, but still having issues with the programming. The code has turned out to be much uglier than i had anticipated, which is one of the main causes of the problems. Also I have talked with Charlie and found out that the parallel environment to use is MPI.

Late October

With the deadline for the Project proposal quickly approaching, I needed to decide which direction my project would need to head. I could stay with CKY and try to answer some questions posed by Hill and Wayne in their paper outlining a parallel CKY algorithm (abandoning all the work I put into learning Earley) or I could try and incorporate some of the techniques used by Hill and Wayne to the parallelization of Earley (quite definitely a much harder task). In the end, Earley won out, and my project is now headed down an unknown and possibly very arduous road.

Mid October

I have continued to improve my survey paper, and have expanded several sections to hopefully better cover what was needed. While undertaking this I realized that I was not very well versed in Earley. Began to concentrate on gaining a better understanding of Earley.

October 4

After meeting with Jim, came to realization that my survey does not cover nearly enough. Need to improve section on CKY, drawing motivation from the inference rules of Schieber. Need to add section for describing various parallel environments.
For presentation, need to better develop slides for parsing. also add slides to aid in the discussion of parallel environments. Also need to add at least one slide on a previous attempt and the methods used.

October 3

Worked on finalizing survey. Finished up the slides for grammars

September 27

Got tiraded by Jim, because I hadn't thought enough about my survey presentation. Continued to work on survey, though my explanation of parsing leaves much to be desired.

September 22

Improved webpage to use PHP to handle including the menus, so i don't have to change each page individually.

September 21

Made the two senior sem webpages you are viewing now.
I used CSS to generate the menus and to format the page. Right now the page only works correctly in mozilla.
Also worked on the survey. Drew upon textbook from Parallel Computations to develop an overview of parallelism.

Previously

Produced the first draft of the Annotated bibliography. I still don't have any sources listed that give a general concept of parsing, nor any sources giving a general concept of parallelism. However I found several papers relating directly to previous attempts at extending CKY and other chart parsers to a parallel environment.