Having finished reviewing the paper on Quantum Computing I started on my research project. Well its supposed to be all finished (since i handed down the thesis a week ago) but i have a couple of weeks to add some finishing touches before the defence... Actually it IS finished and working, but something about the design bothers me. So im sort of re-designing the lexicon structure, in a way that wont affect the thesis!! :)
i need to work on the GUI as well..Re-designing is boooring, since all i need to do is shuffle the contents of the lexicons so that i get ONE core lexicon and no redundencies. So the cerebral effort needed is zilch.
i was going to start coding the GUI when i realized my working code is in the uni lab computer. bummer! so i have to start it tomo, as uni is opening tomo after the Vesak holidays.
so meantime im stuck with shuffling the lexicons, trying to remove the redundant verb roots...
For those of you who are wondering what the heck Im talking about, my research is a parser for Sinhala language. Well, Sinhala Verbs to be exact...Currently it shows 61% precision (or is it accaracy? whatever...If i give it 100 verbs it can identify and parse 61 of them correctly) which would increase as I train it on more data. Im hoping to train it on at least 1000 words (currently its 500) so I might get it up to about 70%.
i need to work on the GUI as well..Re-designing is boooring, since all i need to do is shuffle the contents of the lexicons so that i get ONE core lexicon and no redundencies. So the cerebral effort needed is zilch.
i was going to start coding the GUI when i realized my working code is in the uni lab computer. bummer! so i have to start it tomo, as uni is opening tomo after the Vesak holidays.
so meantime im stuck with shuffling the lexicons, trying to remove the redundant verb roots...
For those of you who are wondering what the heck Im talking about, my research is a parser for Sinhala language. Well, Sinhala Verbs to be exact...Currently it shows 61% precision (or is it accaracy? whatever...If i give it 100 verbs it can identify and parse 61 of them correctly) which would increase as I train it on more data. Im hoping to train it on at least 1000 words (currently its 500) so I might get it up to about 70%.
Comments
Post a Comment