Sunday, September 9, 2012

UDHC- Prediction System

Medical Word Prediction System.

The proposed system takes a passage as an input, extracts some medical words and compares those predicted words with a xml document. The number of match found in the xml document is returned.

The flow chart of the prediction system :-

1. Take input.
2. Extract some medical words using a DataBase.
3. Read the xml document.
4.Search those medical words in the xml document.
5. Count number of match found.
6. Write the output with the passage read from xml document and the no of match found.
7. Xml file is generated.

Pseudo code :


1. Input message= Take Input
2. Predicted word[] = Predict_Words(message,Dictionary)
3. While ( Total No of sample passage in xml document)
4. {
5.      Input passage= Read_Passage_From_Xml_Doc(link1)
6.      count=Search Word (passage, Predicted Word)
7.      Write_OutPut_Xml_File(passage,count);
8. }




Here "" shows the taken passage from a BMJ site.
"" returns number of match found.

Sunday, April 29, 2012

Eclipse Installation in LINUX.


1) Extract the Zip file in a folder. (/home/sumon/Eclipse.)
2) Go to the directory where the Eclipse executable file is exist.
3) ./eclipse
4) If you get an error


http://pastebin.com/UZGXYfWD



5) vi /home/sumon/Eclipse/eclipse.ini

add the line at the end of the file. Before add the line, check if the folder is located in the location /usr/lib/xulrunner-1.9.2/ or not.

-Dorg.eclipse.swt.browser.XULRunnerPath=/usr/lib/xulrunner-1.9.2/

The final eclipse.ini looks like this

http://pastebin.com/siZxD4u5

6) ./eclipse.

I hope the problem will solve.


Friday, April 6, 2012

Basic Use of Sphinx

How to use sphinx4 in Java. This the basic steps to use it.

This is the page where the audio applet is loaded.

This is the page where the audio applet is written.

The applet sends the recorded data to server and server catchs the data and stores into a file. It requests another servlet to convert the audio file into text. Here Sphinx4 is used.
where the recorded data is converted into text send back to the applet.

when i run the page,
i have to press any key from keyboard and have to speak some words according to the grammar file. when i release the space button, it will upload to the server. The server converts the audio stream into the text and sends back to the browser. you can see the output in the page.