Commit Graph

14 Commits

Author SHA1 Message Date
Giambattista Bloisi 8ac167e420 Refinements to PR #404: refactoring the Oaf records merge utilities into dhp-common 2024-04-16 17:18:28 +02:00
Giambattista Bloisi 43b454399f - Bug fix in matchOrderedTokenAndAbbreviations algorithms where tokens with same initial character were always considered equal
- AuthorsMatch exploits the new matching strategy used for ORCID enhancements in #PR398: split author names in tokens, order the tokens, then check for matches of ordered full tokens or abbreviations
2024-04-15 18:19:29 +02:00
Giambattista Bloisi d65285da7f Promote "Research" to a jolly instanceType in dedup comparisons
Compare "Journal" and "Part of book or chapter of book" with "Article"
2024-02-15 12:11:04 +01:00
Giambattista Bloisi 29194472a7 Promote "Research" to a jolly instanceType in dedup comparisons
Compare Part of book or chapter of book with Article
2024-02-15 11:53:46 +01:00
Giambattista Bloisi 02636e802c SparkCreateSimRels:
- Create dedup blocks from the complete queue of records matching cluster key instead of truncating the results
- Clean titles once before clustering and similarity comparisons
- Added support for filtered fields in model
- Added support for sorting List fields in model
- Added new JSONListClustering and numAuthorsTitleSuffixPrefixChain clustering functions
- Added new maxLengthMatch comparator function
- Use reduced complexity Levenshtein with threshold in levensteinTitle
- Use reduced complexity AuthorsMatch with threshold early-quit
- Use incremental Connected Component to decrease comparisons in similarity match in BlockProcessor
- Use new clusterings configuration in Dedup tests

SparkWhitelistSimRels: use left semi join for clarity and performance

SparkCreateMergeRels:
- Use new connected component algorithm that converge faster than Spark GraphX provided algorithm
- Refactored to use Windowing sorting rather than groupBy to reduce memory pressure
- Use historical pivot table to generate singleton rels, merged rels and keep continuity with dedupIds used in the past
- Comparator for pivot record selection now uses "tomorrow" as filler for missing or incorrect date instead of "2000-01-01"
- Changed generation of ids of type dedup_wf_001 to avoid collisions

DedupRecordFactory: use reduceGroups instead of mapGroups to decrease memory pressure
2024-01-10 22:59:52 +01:00
Giambattista Bloisi c412dc162b Fix bug in conversion from dedup json model to Spark Dataset of Rows: list of strings contained the json escaped representation of the value instead of the plain value, this caused instanceTypeMatch failures because of the leading and trailing double quotes 2023-10-02 11:34:51 +02:00
Giambattista Bloisi 3c47920c78 Use asScala to convert java List to Scala Sequence 2023-10-02 11:04:47 +02:00
Giambattista Bloisi e239b81740 Fix defect #8997: GenerateEventsJob is generating huge amounts of logs because broker entity similarity calculation consistently failed 2023-10-02 11:04:18 +02:00
Sandro La Bruzzo 76476cdfb6 Added maven repo for dependencies that are not in maven central 2023-09-20 10:33:14 +02:00
Claudio Atzori bf35280ea6 code formatting 2023-08-29 11:11:00 +02:00
Giambattista Bloisi e64c2854a3 Refactor Dedup process to use Spark Dataframe API and intermediate representation with Row interface
JsonPath cache contention fixed by using a ConcurrentHashMap
Blacklist filtering performance improvement
Minor performance improvements when evaluating similarity
Sorting in clustered elements is deterministic (by ordering and identity field, instead of ordering field only)
2023-07-24 15:36:24 +02:00
Giambattista Bloisi bb5b845e3c Use scala.binary.version property to resolve scala maven dependencies
Ensure consistent usage of maven properties
Profile for compiling with scala 2.12 and Spark 3.4
2023-07-24 11:13:48 +02:00
Giambattista Bloisi 801da2fd4a New sources formatted by maven plugin 2023-07-06 10:28:53 +02:00
Giambattista Bloisi bd3fcf869a rename dnet-pace-core into dhp-pace-core module and use it as dependency in other modules 2023-07-06 10:02:23 +02:00