English
资源说明：Search Engine using inverted index of words with user interface.
Project on Information Retrieval by Kedar Phadtare
Files/folders in this project:
cacm.all contains information about the corpus.
common_words contains common words to be filtered (stop words).
portStemmer.pyc code for stemming of words in corpus.
portStemmer.py Obtained from http://tartarus.org/martin/PorterStemmer/python.txt.
createIndex.py Creates the below mentioned files which form the inverted index for the corpus. (runtime on my machine = 12 secs).
-parseoutputfile1.txt Contains sanitized version of the word in the corpus cacm.all.
-file1 Contains the words of the corpus in the form : [Term-TermID-Offset-CTF-DF].
-file2 Contains the words of the corpus in the form : [TermID-DocID-TermFreq].
-file3 Contains the words of the corpus in the form : [DocID-DocName-Doclen].
-newindex.txt Contains the actual inverted index for the words in the corpus.
webinterface.html Code for web page with text box for entering query.
userokbm25.py Code for processing user entered query and displaying ranked list of documents using Okapi BM25 retrieval model.
Shows the first 10 most relevant document names as links.
snippet.py Code to display the actual data in the document using the click on the link of the document.
Locations:
The folder named 'files' should be present in the directory where createIndex.py is to be run. This is because to recude significant computing time,
individual files are created for all the unique words in the corpus and their information is just updated in these files.
The webinterface.html,userokbm25.py,snippet.py,common_words,porterStemmer files along with the generated files (file1,file2,newindex.txt,parseoutputfile1.txt) should be kept in the /usr/lib/cgi-bin folder on a machine to be able to view via browser.
The Inverted Index can be created anywhere on a machine.
本源码包内暂不包含可直接显示的源代码文件，请下载源码包。
