This is a command-line based Inforamtion Retrieval System built in JAVA using Apache Lucene search library 1.
- Using Lucene, it parses and indexs Plain Text and HTML documents that a given folder and its subfolders contain. Also, all parsed files are listed.
- English language is considered to Stem (e.g. using Porter Stemmer).
- A search index is selected or a new one is created (if not available in the chosen directory).
- User can choose the ranking model with keywords as; Vector Space Model (VS) or Okapi BM25 (OK).
- A ranked list of relevant articles given a search query are printed. The output contains the 10 most relevant documents with their rank, relevance score, file name resp. title (if it is a HTML-file containing title information) and the path.
- Multiple fields are concurrently searched (multifield search): not only the document’s text (body tag) are searched, but also its title.
To know more about what an individual JAVA files do, please refer the documentation.
Step1: Keep jar file in your desired location.
Step2: Make sure that you have created index directory and document directory which consist of sample files for indexing in desired location.
Step3: Open command line and change directory to the directory where jar file has been present.
Step4: Enter the command as:
java -jar IR_P.jar [path to document folder] [path to index folder] [VS/OK] [query]
Step5: Results will be displayed on the console.