README.md

Scoring and Relevancy example

This comics example is mainly about relevancy and scoring.

Redis University Link

• Redis University Link
• text transcript file included from the redis university training as well
• comics web site inspiring the dataset
• Create Index

redis-cli < create_idx.txt

• Add two non-weighted records

redis-cli < add_comics.txt

• search using keys hero

redis-cli < queries.txt

The resulting relevancy (score) tells us that Spiderman is a more likely hero than Batman because the word 'hero' appears more often for 'Spiderman' (term frequency).

• modify the data to add a rating column and add superman

redis-cli < modify.txt

• Run same search again now with ratings considered

redis-cli < queries.txt

now adding weights

redis-cli flushall
redis-cli < create_idx_rating.txt
redis-cli < add-comics2.txt
redis-cli < queries.txt

• run explainscore for more detail

redis-cli < queries_explain.txt