1.0.16

• Underscores do not count as word characters in the tokenizer (e.g., "retinoic_acid" becomes "retinoic acid" in the full text search)

1.0.15

• Added support for optional chi-square test

1.0.14

• Added support for multiple neo4j containers

1.0.13

• Fixed a bug in parsing abstract titles when there was an italicized word in the title
• Fixed a bug when a query included "x&y" and x or y were in other terms of the query

1.0.12

• fixed pubmed cases where year was not parsed correctly

1.0.11

• Fixed an issue where B-terms with no significant C-term pairs would crowd out other B-terms if their A-B prediction score was high

1.0.10

• Knowledge graph querying is faster
• Fixed small calculation error in the prediction score
• Much improved RAM efficiency
• Implemented different priority queues
• Implement synonyms and date censoring for knowledge graph
• PMIDs are ordered by citation count

1.0.9

• Add endpoint to cancel a job
• Add neo4j knowledge graph querying

1.0.8

• Fixed a bug in reporting percent progress complete

1.0.7

• Job IDs can be provided when submitting a job, instead of having one always be randomly generated