mattszeto / findex Goto Github PK

Search Engine:

• Boolean Queries: Positional Inverted Index, K-gram Index, Porter2Stemmer Algorithm
  • K-Grams
  • Stemming
  • NEAR/k Queries
  • Wildcard Queries
• Ranked Queries: Disk Positional Index, K-gram Index on Disk
  • Spelling Check (Suggested Queries): Jaccard Coefficient, Levenshtein Edit Distance

Dependencies:

• SparkJava w/ Thymeleaf-Template: com.sparkjava:spark-template-thymeleaf:2.7.1
• Gson: gson-2.8.6.jar
• libstemmer (porter2stemmer): tartarus snowball-stemmer
• mapdb (b+tree for on-disk index): org.mapdb

You can run our search engine in the command line by running indexer.java

1. Open up your favorite text-editor or IDE (project developed on intellij)
2. download and load all dependencies (noted above)
3. run indexer.java for a fully functional cmd text application

Another way is to run from your local machine at http://localhost:4567/

1. Open up text-editor or IDE (project developed on intellij)
2. download and load all dependencies (noted above)
3. run WebUI.java
4. open a chrome window and go to http://localhost:4567/

To build an index, provide a path to your corpus folder.

(if you have already built your index before, and no files have changed within the corpus, then you do not need to build index again) you can check this by seeing if your corpus folder already has an index folder inside with postings.bin and docWeights.bin (there will be other files too, but these are the most important

You must bring your own corpus to index and search from (local file directory needed)