52 lines
2.3 KiB
Plaintext
52 lines
2.3 KiB
Plaintext
{
|
|
"wf" : {
|
|
"threshold" : "0.99",
|
|
"dedupRun" : "001",
|
|
"entityType" : "result",
|
|
"orderField" : "title",
|
|
"queueMaxSize" : "2000",
|
|
"groupMaxSize" : "10",
|
|
"slidingWindowSize" : "200",
|
|
"rootBuilder" : [ "result" ],
|
|
"includeChildren" : "true"
|
|
},
|
|
"pace" : {
|
|
"clustering" : [
|
|
{ "name" : "acronyms", "fields" : [ "title" ], "params" : { "max" : "1", "minLen" : "2", "maxLen" : "4"} },
|
|
{ "name" : "ngrampairs", "fields" : [ "title" ], "params" : { "max" : "1", "ngramLen" : "3"} },
|
|
{ "name" : "suffixprefix", "fields" : [ "title" ], "params" : { "max" : "1", "len" : "3" } }
|
|
],
|
|
"necessaryConditions" : [
|
|
{ "name" : "yearMatch", "fields" : [ "dateofacceptance" ] },
|
|
{ "name" : "titleVersionMatch", "fields" : [ "title" ] },
|
|
{ "name" : "sizeMatch", "fields" : [ "authors" ] } ,
|
|
{ "name" : "pidMatch", "fields" : [ "pid" ] }
|
|
],
|
|
"model" : [
|
|
{ "name" : "pid", "algo" : "Null", "type" : "JSON", "weight" : "0.0", "ignoreMissing" : "true", "path" : "pid", "overrideMatch" : "true" },
|
|
{ "name" : "title", "algo" : "JaroWinkler", "type" : "String", "weight" : "1.0", "ignoreMissing" : "false", "path" : "result/metadata/title[qualifier#classid = {main title}]/value" },
|
|
{ "name" : "dateofacceptance", "algo" : "Null", "type" : "String", "weight" : "0.0", "ignoreMissing" : "true", "path" : "result/metadata/dateofacceptance/value" } ,
|
|
{ "name" : "authors", "algo" : "Null", "type" : "String", "weight" : "0.0", "ignoreMissing" : "true", "path" : "result/metadata/author/fullname" }
|
|
],
|
|
"blacklists" : {
|
|
"title" : [
|
|
"^(Corpus Oral Dialectal \\(COD\\)\\.).*$",
|
|
"^(Kiri Karl Morgensternile).*$",
|
|
"^(\\[Eksliibris Aleksandr).*\\]$",
|
|
"^(\\[Eksliibris Aleksandr).*$",
|
|
"^(Eksliibris Aleksandr).*$",
|
|
"^(Kiri A\\. de Vignolles).*$",
|
|
"^(2 kirja Karl Morgensternile).*$",
|
|
"^(Pirita kloostri idaosa arheoloogilised).*$",
|
|
"^(Kiri tundmatule).*$",
|
|
"^(Kiri Jenaer Allgemeine Literaturzeitung toimetusele).*$",
|
|
"^(Eksliibris Nikolai Birukovile).*$",
|
|
"^(Eksliibris Nikolai Issakovile).*$",
|
|
"^(WHP Cruise Summary Information of section).*$",
|
|
"^(Measurement of the top quark\\-pair production cross section with ATLAS in pp collisions at).*$",
|
|
"^(Measurement of the spin\\-dependent structure function).*"
|
|
] }
|
|
}
|
|
|
|
}
|