xuechaofelix / textmessageclassification Goto Github PK

What is the file meaning?
/Divide : 
	it is the eclipse project and it can be imported by eclipse. its function is Word segmentation.
/Divide/IKAnalyzer2012_u6.jar:
	the tool jar for word segmentation
/Divide/lucene-core-3.6.0.jar:
	the tool jar for word segmentation
/Divide/*_result.txt
	the output file after word segmentation
/Message.py : 
	get the feature from the training text by IG. 
/all_feature.txt: 
	The output of the Message.py with all the feature of this text.
/deal.py: 
	read the feature and training the training data. Print the Right-Rate and get testing_classify-result.
/feature.txt: 
	the feature whose distinction is biggger.
/testing_classify_result.txt: 
	the output of the deal.py with the classify result for testing file
/testing_result.txt: 
	The output of the Divide project with all the possiable word in the testing file.
/training_result.txt: 
	The output of the Divide project with all the possiable word in the training file.


Worning: 
	Default Running Enviroment Is Python 3.6, If You Do Not Have It, Please Delete The First Line In The Python File. i.e. (Del: #! /usr/bin/env python3.6)


Waht is the Runing Order?

Note: the every following step is the independent step. You can run any step you want.

1.Word Segmentation
 >Import the eclipse, and open the test.java. 
 >Change the paramter of the FileReader to the name of which text you want to do word segmentation. 
 >Change the paramter of the FileWrite to the file name you want to output. 
 >Run 

2.Get Feature
 >Change all the Input and Output file name which is you want to run
 >Change how many features you want get by the NumOfFeature
 >Run message.py

3.Get the Result
 >Change all the Input and Output file name which is you want to run 
 >Run deal.py

What is the result of the Running?

>Word Segmentation
	It get a file which is divided by word segmentation with keeping the original file form.
>Get Feature
	Get two files. one are the featrues we want to get with the form of ('word','distinction'). another are all the features of the file with the form of ('word','distinction').
>Get the Result
	Print the Right-Rate
	Get a file which is the testing result with the form 0 or 1 following the roiginal sentence for each line of the tesing file.