http://blogs.msdn.com/b/webngram/archive/2010/11/08/using-the-microsoftngram-python-module.aspxA quick tutorial on the MicrosoftNgram Python library.en-USTelligent Evolution Platform Developer Build (Build: 5.6.50428.7875)http://blogs.msdn.com/b/webngram/archive/2010/11/08/using-the-microsoftngram-python-module.aspx#10097411Sat, 27 Nov 2010 21:09:54 GMT91d46819-8472-40ad-a661-2c78acb4018c:10097411Viswanath<p>We have one problem: We need all 5-grams which have a particular word in them (Say for example: All 5-grams which have sachin in it). Currently with generate service we can only get &nbsp;all &quot;one succeeding&quot; word after a given context(phrase). Is there any way so that we can get all 5-grams which have a particular word in them.</p> <p>We figure out this way:</p> <p> &nbsp; &nbsp; &nbsp;1. First call the generative service: through which we will get all one succeeding words.</p> <p> &nbsp; &nbsp; &nbsp;2. Now we have a context say &quot;sachin tendulkar&quot; and then we make a one more Generative Service call through which we will get all &quot;one succeeding&quot; words after context. And we will multiple both &nbsp; &nbsp; probabilities ( Joint Probability ).</p> <p>Now the problem is: Through this procedure we can only get succeeding words, we need left context of a given particular word. If you have any solution, please let us know. </p> <div style="clear:both;"></div><img src="http://blogs.msdn.com/aggbug.aspx?PostID=10097411" width="1" height="1">