release 2024-08-09

- add supersenses to WN-LMF export - include marl as export prefix for DanNet dataset (new sentiment data)
kuhumcst · Aug 9, 2024 · 63fbc81 · 63fbc81
1 parent ff90adc
commit 63fbc81
Show file tree

Hide file tree

Showing 6 changed files with 24 additions and 9 deletions.
diff --git a/examples/wn_lmf_query.py b/examples/wn_lmf_query.py
@@ -3,6 +3,6 @@
 # based on the example from the https://github.com/goodmami/wn README
 # (to clear db between restarts: rm ~/.wn_data/wn.db)
 
-wn.add("../export/wn-lmf/dannet-wn-lmf.gz")
+wn.add("../export/wn-lmf/dannet-wn-lmf.xml.gz")
 ss = wn.synsets('vinde', pos='v')[0]
-print(ss.definition())
+print(ss.definition() + " (" + ss.lexfile() + ")")
diff --git a/pages/releases-da.md b/pages/releases-da.md
@@ -1,7 +1,7 @@
 # Versioner
 De nye DanNet-versioner bruger udgivelsesdatoen som versionsnummer, formateret som `YYYY-MM-DD`.
 
-## **SNAPSHOT**: Supersenses + OEWN-opdatering
+## **2024-08-09**: Supersenses + OEWN-opdatering
 * Tilføjede 71055 Supersenses til DanNet, primært baseret på en mapping skabt til SemDaX-korporaet.
   * Omkring 300 af disse er "rettet" efter hypernym til forskel fra mappingen.
   * Yderligere 3793 af disse er blevet til på baggrund af en ny mapping af ikke-mappede synsets.

diff --git a/pages/releases-en.md b/pages/releases-en.md
@@ -1,7 +1,7 @@
 # Releases
 The newer DanNet releases use the release date as the version number, formatted as `YYYY-MM-DD`.
 
-## **SNAPSHOT**: Supersenses + OEWN update
+## **2024-08-09**: Supersenses + OEWN update
 * Added 71055 Supersenses to DanNet, mostly based on a mapping devised for the SemDaX corpora.
   * Around 300 of these were "fixed" according to hypernym, differing from the mapping.
   * Another 3793 of these were created based on a new mapping of unmapped synsets.

diff --git a/src/main/dk/cst/dannet/db/bootstrap.clj b/src/main/dk/cst/dannet/db/bootstrap.clj
@@ -91,7 +91,7 @@
   "2024-06-12")
 
 (def current-release
-  (str "2024-06-12" "-SNAPSHOT"))
+  (str "2024-08-09" #_"-SNAPSHOT"))
 
 (defn assert-expected-dannet-release!
   "Assert that the DanNet `model` is the expected release to boostrap from."
@@ -351,7 +351,7 @@
   This function survives between releases, but the functions it calls are all
   considered temporary and should be deleted when the release comes."
   [dataset]
-  (let [expected-release "2024-06-12-SNAPSHOT"]
+  (let [expected-release "2024-08-09"]
     (assert (= current-release expected-release))           ; another check
     (println "Applying release changes for" expected-release "...")
 

diff --git a/src/main/dk/cst/dannet/db/export/wn_lmf.clj b/src/main/dk/cst/dannet/db/export/wn_lmf.clj
@@ -138,6 +138,13 @@
        ?synset skos:definition ?definition .
      }"))
 
+(def supersense-query
+  (op/sparql
+    "SELECT ?synset ?supersense
+     WHERE {
+       ?synset dns:supersense ?supersense .
+     }"))
+
 (def pos-str
   {:wn/adjective "a"
    :wn/noun      "n"
@@ -163,9 +170,10 @@
 
 (defn synset
   [synset-props [id ms]]
-  (let [{:keys [ili pos definition examples members]} (get synset-props id)]
+  (let [{:keys [ili pos definition lexfile examples members]} (get synset-props id)]
     (into [:Synset (cond-> {:id           (name id)
                             :members      (str/join " " members)
+                            :lexfile      lexfile
                             :ili          (or ili "")       ; attr required by https://github.com/goodmami/wn
                             :partOfSpeech (or pos "")})     ; attr required by https://github.com/goodmami/wn
            (when definition
@@ -239,6 +247,9 @@
         definition-query-res  (label-time
                                 'definition-query-res
                                 (q/run g definition-query))
+        supersense-query-res  (label-time
+                                'supersense-query-res
+                                (q/run g supersense-query))
         get-relations-res     (label-time
                                 'get-supported-relations
                                 (get-supported-relations g))
@@ -278,7 +289,11 @@
                  (update-vals
                    (group-by '?synset definition-query-res)
                    (fn [ms]
-                     {:definition (-> ms first (get '?definition) str)})))]))
+                     {:definition (-> ms first (get '?definition) str)}))
+                 (update-vals
+                   (group-by '?synset supersense-query-res)
+                   (fn [ms]
+                     {:lexfile (-> ms first (get '?supersense))})))]))
 
 (defn xml-str
   "Create a valid WN-LMF XML string from `query-results`."

diff --git a/src/main/dk/cst/dannet/prefix.cljc b/src/main/dk/cst/dannet/prefix.cljc
@@ -90,7 +90,7 @@
              :resource (str "<" dannet-root "data>")
              :export   #{'dn 'dnc 'dns
                          'rdf 'rdfs 'owl
-                         'wn 'ontolex 'skos 'lexinfo
+                         'wn 'ontolex 'skos 'lexinfo 'marl
                          'dcat 'vann 'foaf 'dc
                          'ili 'en}
              :download {"rdf"    {:default   "dannet.zip"