| /* |
| * Description : Fuzzy self joins a dataset, DBLP, based on the similarity-jaccard function of its titles' word tokens. |
| * DBLP has a keyword index on title, and we expect the join to be transformed into an indexed nested-loop join. |
| * We test the inlining of variables that enable the select to be pushed into the join for subsequent optimization with an index. |
| * Success : Yes |
| */ |
| |
| drop dataverse test if exists; |
| create dataverse test; |
| use dataverse test; |
| |
| create type DBLPType as closed { |
| id: int32, |
| dblpid: string, |
| title: string, |
| authors: string, |
| misc: string |
| } |
| |
| create dataset DBLP(DBLPType) partitioned by key id; |
| |
| load dataset DBLP |
| using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter" |
| (("path"="nc1://data/dblp-small/dblp-small-id.txt"),("format"="delimited-text"),("delimiter"=":")) pre-sorted; |
| |
| create index keyword_index on DBLP(title) type keyword; |
| |
| write output to nc1:"rttest/inverted-index-join_word-jaccard_04.adm"; |
| |
| for $a in dataset('DBLP') |
| for $b in dataset('DBLP') |
| let $jacc := similarity-jaccard(word-tokens($a.title), word-tokens($b.title)) |
| where $jacc >= 0.5f and $a.id < $b.id |
| return {"arec": $a, "brec": $b, "jacc": $jacc } |