As you have noticed there are some work done in the field of Sinhala
spell checking and support for popular applications such as Firefox,
OpenOffice.etc. Currently most of these applications are using Hunspell
<http://hunspell.sourceforge.net/> as the framework for spell checking
and using a common framework across many applications comes handy
because it gives chance to focus on improving wordlists rather than
getting used to various technologies.
At the moment there are several wordlists including UCSC word list
<http://ucsc.cmb.ac.lk/ltrl/downloader.php?resource=wdlst> and one
<https://github.com/buddhika/Sinhala-Dictionary> being developed by me.
If you are interested in this approach of building/improving spell
checkers it might be better to setup Hunspell and check how it can
workout with Sinhala (and this
<http://www.lakapps.lk/clsa/slides-10/4_3.pdf> might help as a starting
point). Another approach would be to forgo any frameworks and just build
a spell checker from the scratch specifically for Sinhala. This could be
advantageous when exploiting some unique Sinhala language features such
as some forms of "Sandhi" to make spell checking process more corrective
but on the other hand will have the disadvantage of being unusable with
common text editing applications.
Hope this helps.
On 01/01/2011 06:08 PM, Maheeka Jayasuriya wrote:
> Thank you for your concern regarding this matter. I hope to get help
> and support from all experienced and knowledged in this area as my
> project is continuing. Thanks
> On Sat, Jan 1, 2011 at 4:49 PM, Harshula <harshula@...
> <mailto:harshula@...>> wrote:
> Hi All,
> There's a number of you that can probably help Maheeka.
> -------- Forwarded Message --------
> I am a final year student at APIIT. I am to select my final year
> project within this week. After going through so many ideas, i have
> ended up with a few useful ideas regarding Sinhala.
> Initially i wanted to create a translator for English to Sinhala,
> which i know is not an easy task. And it is almost impossible in 8
> months of my FYP period. So after talking with my supervisor, he
> advised me to go for a Sinhala to English translator, which is a
> little bit more manageable than vice-versa, considering the time. Even
> out of this i had to select a limited area of the language. The reason
> I had to limit my scope is that I have to develop everything from
> scratch. Therefore, I have selected sentences adhering to "prathama
> wibhakthi" and "sambanda wibhakthi" only.
> My other idea was to create a Sinhala spell checker. After some
> googling i found out that there is already a spell checker for open-
> office and an add-on for Firefox. the firefox add-on does not support
> with the newer versions and the open office plug-in needs some
> improvements regarding the word set. After going through the threads
> in this forum, i saw a thread that was referring to the need of a
> browser spell-checker. I further explored this idea and informed my
> supervisor of my intention to develop such a plug-in. I personally
> prefer this over the translator. But I have not confirmed my selection
> as at yet.
> Since it is the vacation I still couldn't get my supervisor's opinion
> about these two ideas. But since both is regarding Sinhala, i would
> like to get support from this group members for this project. i would
> like to hear your ideas regarding this matter.