DokumentationExperience PlatformHandbuch zum Abfragedienst

Ingenieurfunktionen für maschinelles Lernen

Last update: Mon Dec 30 2024 00:00:00 GMT+0000 (Coordinated Universal Time)

Erstellt für:

  • Benutzende
  • Entwickler

In diesem Dokument wird gezeigt, wie Sie Daten in Adobe Experience Platform in Funktionen oder Variablen umwandeln können, die von einem maschinellen Lernmodell genutzt werden können. Dieser Prozess wird als „Feature ". Verwenden Sie Data Distiller, um ML-Funktionen im benötigten Umfang zu berechnen und diese Funktionen für Ihre maschinelle Lernumgebung freizugeben. Dies umfasst Folgendes:

  1. Erstellen Sie eine Abfragevorlage, um die Zielbeschriftungen und -funktionen zu definieren, die Sie für Ihr Modell berechnen möchten
  2. Ausführen der Abfrage und Speichern der Ergebnisse in einem Trainings-Datensatz

Definieren von Schulungsdaten define-training-data

Das folgende Beispiel zeigt eine Abfrage zum Ableiten von Schulungsdaten aus einem Erlebnisereignis-Datensatz für ein Modell, um die Neigung eines Benutzers vorherzusagen, einen Newsletter zu abonnieren. Abonnementereignisse werden durch den Ereignistyp web.formFilledOut dargestellt. Andere Verhaltensereignisse im Datensatz werden verwendet, um Funktionen auf Profilebene abzuleiten und Abonnements vorherzusagen.

Abfrage von positiven und negativen Kennzeichnungen query-positive-and-negative-labels

Ein vollständiger Datensatz für das Training eines (überwachten) Modells für maschinelles Lernen enthält eine Zielvariable oder -bezeichnung, die das vorherzusagende Ergebnis darstellt, und einen Satz von Funktionen oder erklärenden Variablen, die zur Beschreibung der Beispielprofile verwendet werden, die zum Trainieren des Modells verwendet werden.

In diesem Fall ist die Bezeichnung eine Variable mit der Bezeichnung subscriptionOccurred , die gleich 1 ist, wenn das Benutzerprofil ein Ereignis mit dem Typ web.formFilledOut hat, andernfalls ist es 0. Die folgende Abfrage gibt einen Satz von 50.000 Benutzern aus dem Ereignisdatensatz zurück, einschließlich aller Benutzer mit positiven Kennzeichnungen (subscriptionOccurred = 1) plus einem Satz zufällig ausgewählter Benutzer mit negativen Kennzeichnungen, um die Stichprobengröße von 50.000 Benutzern abzuschließen. Dadurch wird sichergestellt, dass die Trainings-Daten sowohl positive als auch negative Beispiele für das Modell enthalten, aus denen gelernt werden kann.

from aepp import queryservice

dd_conn = queryservice.QueryService().connection()
dd_cursor = queryservice.InteractiveQuery2(dd_conn)

query_labels = f"""
SELECT *
FROM (
    SELECT
        eventType,
        _{tenant_id}.user_id as userId,
        SUM(CASE WHEN eventType='web.formFilledOut' THEN 1 ELSE 0 END)
            OVER (PARTITION BY _{tenant_id}.user_id)
            AS "subscriptionOccurred",
        row_number() OVER (PARTITION BY _{tenant_id}.user_id ORDER BY randn()) AS random_row_number_for_user
    FROM {table_name}
)
WHERE (subscriptionOccurred = 1 AND eventType = 'web.formFilledOut') OR (subscriptionOccurred = 0 AND random_row_number_for_user = 1)
"""

df_labels = dd_cursor.query(query_labels, output="dataframe")
print(f"Number of classes: {len(df_labels)}")
df_labels.head()

Beispielausgabe

Anzahl Klassen: 50000

eventType
userId
subscriptionOccurred
random_row_number_for_user
0
directMarketing.emailClicked
01027994177972439148069092698714414382
0
1
1
directMarketing.emailOpened
01054714817856066632264746967668888198
0
1
2
web.formFilledOut
01117296890525140996735553609305695042
1
15
3
directMarketing.emailClicked
01149554820363915324573708359099551093
0
1
4
directMarketing.emailClicked
01172121447143590196349410086995740317
0
1

Aggregieren von Ereignissen zur Definition von Funktionen für ML define-features

Mit einer entsprechenden Abfrage können Sie die Ereignisse im Datensatz in aussagekräftigen numerischen Funktionen erfassen, die zum Trainieren eines Tendenzmodells verwendet werden können. Beispielereignisse werden unten angezeigt:

  • Anzahl der E Mails, die zu Marketing-Zwecken gesendet und vom Benutzer empfangen wurden.
  • Teil dieser E-Mails, die geöffnet wurden.
  • Teil dieser E-Mails, bei dem der Benutzer ausgewählt den Link.
  • Anzahl der Produkte die angezeigt wurden.
  • Anzahl der Vorschläge, mit denen interagiert wurde.
  • Anzahl abgelehnten Vorschläge.
  • Anzahl ausgewählten Links.
  • Anzahl der Minuten zwischen zwei aufeinander folgenden empfangenen E-Mails.
  • Anzahl der Minuten zwischen zwei aufeinander folgenden geöffneten E-Mails
  • Anzahl der Minuten zwischen zwei aufeinander folgenden E-Mails, in denen der Benutzer den Link tatsächlich ausgewählt hat.
  • Anzahl der Minuten zwischen zwei aufeinander folgenden Produktansichten.
  • Anzahl der Minuten zwischen zwei Vorschlägen, mit denen interagiert wurde.
  • Anzahl der Minuten zwischen zwei abgelehnten Vorschlägen.
  • Anzahl der Minuten zwischen zwei ausgewählten Links.

Die folgende Abfrage aggregiert diese Ereignisse:

Auswählen, um Beispielabfrage anzuzeigen
code language-python 
query_features = f"""
SELECT
    _{tenant_id}.user_id as userId,
    SUM(CASE WHEN eventType='directMarketing.emailSent' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "emailsReceived",
    SUM(CASE WHEN eventType='directMarketing.emailOpened' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "emailsOpened",
    SUM(CASE WHEN eventType='directMarketing.emailClicked' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "emailsClicked",
    SUM(CASE WHEN eventType='commerce.productViews' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "productsViewed",
    SUM(CASE WHEN eventType='decisioning.propositionInteract' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "propositionInteracts",
    SUM(CASE WHEN eventType='decisioning.propositionDismiss' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "propositionDismissed",
    SUM(CASE WHEN eventType='web.webinteraction.linkClicks' THEN 1 ELSE 0 END)
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "webLinkClicks" ,
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailSent', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_emailSent",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailOpened', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_emailOpened",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailClicked', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_emailClick",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'commerce.productViews', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_productView",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'decisioning.propositionInteract', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_propositionInteract",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'propositionDismiss', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_propositionDismiss",
    TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'web.webinteraction.linkClicks', 'minutes')
       OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
       AS "minutes_since_linkClick"
FROM {table_name}
"""

df_features = dd_cursor.query(query_features, output="dataframe")
df_features.head()

Beispielausgabe

userId
EmailsReceived
emailsOpened
E-Mails angeklickt
productsViewed
propositionInteractions
Vorschlag abgelehnt
webLinkClicks
minutes_since_emailSent
minutes_since_emailOpened
minutes_since_emailClick
minutes_since_productView
minutes_since_propositionInteract
minutes_since_propositionDismiss
minutes_since_linkClick
0
01102546977582484968046916668339306826
1
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
1
01102546977582484968046916668339306826
2
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
2
01102546977582484968046916668339306826
3
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
3
01102546977582484968046916668339306826
3
1
0
0
0
0
0
540,0
0,0
NaN
NaN
NaN
Keine
NaN
4
01102546977582484968046916668339306826
3
2
0
0
0
0
0
588,0
0,0
NaN
NaN
NaN
Keine
NaN

Kombinieren von Kennzeichnungen und Funktionen in Abfragen combine-queries

Schließlich können die Abfragebeschriftungen und die Abfragefeatures zu einer einzigen Abfrage kombiniert werden, die einen Trainings-Datensatz mit Beschriftungen und Funktionen zurückgibt:

Auswählen, um Beispielabfrage anzuzeigen
code language-python 
query_training_set = f"""
SELECT *
FROM (
    SELECT _{tenant_id}.user_id as userId,
       eventType,
       timestamp,
       SUM(CASE WHEN eventType='web.formFilledOut' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id)
           AS "subscriptionOccurred",
       SUM(CASE WHEN eventType='directMarketing.emailSent' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsReceived",
       SUM(CASE WHEN eventType='directMarketing.emailOpened' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsOpened",
       SUM(CASE WHEN eventType='directMarketing.emailClicked' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsClicked",
       SUM(CASE WHEN eventType='commerce.productViews' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "productsViewed",
       SUM(CASE WHEN eventType='decisioning.propositionInteract' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionInteracts",
       SUM(CASE WHEN eventType='decisioning.propositionDismiss' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionDismissed",
       SUM(CASE WHEN eventType='web.webinteraction.linkClicks' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "webLinkClicks" ,
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailSent', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailSent",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailOpened', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailOpened",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailClicked', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailClick",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'commerce.productViews', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_productView",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'decisioning.propositionInteract', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionInteract",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'propositionDismiss', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionDismiss",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'web.webinteraction.linkClicks', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_linkClick",
        row_number() OVER (PARTITION BY _{tenant_id}.user_id ORDER BY randn()) AS random_row_number_for_user
    FROM {table_name} LIMIT 1000
)
WHERE (subscriptionOccurred = 1 AND eventType = 'web.formFilledOut') OR (subscriptionOccurred = 0 AND random_row_number_for_user = 1)
ORDER BY timestamp
"""

df_training_set = dd_cursor.query(query_training_set, output="dataframe")
df_training_set.head()

Beispielausgabe

userId
eventType
Zeitstempel
subscriptionOccurred
EmailsReceived
emailsOpened
E-Mails angeklickt
productsViewed
propositionInteractions
Vorschlag abgelehnt
webLinkClicks
minutes_since_emailSent
minutes_since_emailOpened
minutes_since_emailClick
minutes_since_productView
minutes_since_propositionInteract
minutes_since_propositionDismiss
minutes_since_linkClick
random_row_number_for_user
0
02554909162592418347780983091131567290
directMarketing.emailSent
17.06.2023 13:44:59.086
0
2
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
1
1
01130334080340815140184601481559659945
directMarketing.emailOpened
19.06.2023:01:06.55.366
0
1
3
0
1
0
0
0
1921,0
0,0
NaN
1703,0
NaN
Keine
NaN
1
2
01708961660028351393477273586554010192
web.formFilledOut
19.06.2023 18:36:49.083
1
1
2
2
0
0
0
0
2365,0
26,0
1,0
NaN
NaN
Keine
NaN
7
3
01809182902320674899156240602124740853
directMarketing.emailSent
21.06.2023 19:17:12.535
0
1
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
1
4
03441761949943678951106193028739001197
directMarketing.emailSent
21.06.2023 21:58:29.482
0
1
0
0
0
0
0
0
0,0
NaN
NaN
NaN
NaN
Keine
NaN
1

Erstellen einer Abfragevorlage zur inkrementellen Berechnung von Schulungsdaten

Es ist üblich, ein Modell regelmäßig mit aktualisierten Trainingsdaten neu zu trainieren, um die Genauigkeit des Modells über die Zeit hinweg aufrechtzuerhalten. Als Best Practice zur effizienten Aktualisierung Ihres Trainings-Datensatzes können Sie eine Vorlage aus Ihrer Trainings-Set-Abfrage erstellen, um neue Trainings-Daten inkrementell zu berechnen. Auf diese Weise können Sie Beschriftungen und Funktionen nur aus Daten berechnen, die seit der letzten Aktualisierung der Schulungsdaten zum ursprünglichen Erlebnisereignis-Datensatz hinzugefügt wurden, und die neuen Beschriftungen und Funktionen in den vorhandenen Schulungsdatensatz einfügen.

Dies erfordert einige Änderungen an der Trainings-Set-Abfrage:

  • Fügen Sie Logik hinzu, um einen neuen Trainings-Datensatz zu erstellen, falls er nicht vorhanden ist, und fügen Sie andernfalls die neuen Beschriftungen und Funktionen in den vorhandenen Trainings-Datensatz ein. Dies erfordert eine Reihe von zwei Versionen der Trainings-Set-Abfrage:

    • Verwenden Sie zunächst die Anweisung CREATE TABLE IF NOT EXISTS {table_name} AS .
    • Verwenden Sie anschließend die INSERT INTO {table_name} für den Fall, dass der Trainings-Datensatz bereits vorhanden ist

  • Fügen Sie eine SNAPSHOT BETWEEN $from_snapshot_id AND $to_snapshot_id Anweisung hinzu, um die Abfrage auf Ereignisdaten zu beschränken, die innerhalb eines bestimmten Intervalls hinzugefügt wurden. Das $ Präfix für die Momentaufnahme-IDs gibt an, dass es sich um Variablen handelt, die bei der Ausführung der Abfragevorlage übergeben werden.

Das Anwenden dieser Änderungen führt zur folgenden Abfrage:

Auswählen, um Beispielabfrage anzuzeigen
code language-python 
ctas_table_name = "propensity_training_set"

query_training_set_template = f"""
$$ BEGIN

SET @my_table_exists = SELECT table_exists('{ctas_table_name}');

CREATE TABLE IF NOT EXISTS {ctas_table_name} AS
SELECT *
FROM (
    SELECT _{tenant_id}.user_id as userId,
       eventType,
       timestamp,
       SUM(CASE WHEN eventType='web.formFilledOut' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id)
           AS "subscriptionOccurred",
       SUM(CASE WHEN eventType='directMarketing.emailSent' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsReceived",
       SUM(CASE WHEN eventType='directMarketing.emailOpened' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsOpened",
       SUM(CASE WHEN eventType='directMarketing.emailClicked' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsClicked",
       SUM(CASE WHEN eventType='commerce.productViews' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "productsViewed",
       SUM(CASE WHEN eventType='decisioning.propositionInteract' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionInteracts",
       SUM(CASE WHEN eventType='decisioning.propositionDismiss' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionDismissed",
       SUM(CASE WHEN eventType='web.webinteraction.linkClicks' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "webLinkClicks" ,
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailSent', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailSent",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailOpened', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailOpened",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailClicked', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailClick",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'commerce.productViews', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_productView",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'decisioning.propositionInteract', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionInteract",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'propositionDismiss', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionDismiss",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'web.webinteraction.linkClicks', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_linkClick",
        row_number() OVER (PARTITION BY _{tenant_id}.user_id ORDER BY randn()) AS random_row_number_for_user
    FROM {table_name}
    SNAPSHOT BETWEEN $from_snapshot_id AND $to_snapshot_id
)
WHERE (subscriptionOccurred = 1 AND eventType = 'web.formFilledOut') OR (subscriptionOccurred = 0 AND random_row_number_for_user = 1)
ORDER BY timestamp;

INSERT INTO {ctas_table_name}
SELECT *
FROM (
    SELECT _{tenant_id}.user_id as userId,
       eventType,
       timestamp,
       SUM(CASE WHEN eventType='web.formFilledOut' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id)
           AS "subscriptionOccurred",
       SUM(CASE WHEN eventType='directMarketing.emailSent' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsReceived",
       SUM(CASE WHEN eventType='directMarketing.emailOpened' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsOpened",
       SUM(CASE WHEN eventType='directMarketing.emailClicked' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "emailsClicked",
       SUM(CASE WHEN eventType='commerce.productViews' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "productsViewed",
       SUM(CASE WHEN eventType='decisioning.propositionInteract' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionInteracts",
       SUM(CASE WHEN eventType='decisioning.propositionDismiss' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "propositionDismissed",
       SUM(CASE WHEN eventType='web.webinteraction.linkClicks' THEN 1 ELSE 0 END)
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "webLinkClicks" ,
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailSent', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailSent",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailOpened', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailOpened",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'directMarketing.emailClicked', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_emailClick",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'commerce.productViews', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_productView",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'decisioning.propositionInteract', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionInteract",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'propositionDismiss', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_propositionDismiss",
       TIME_BETWEEN_PREVIOUS_MATCH(timestamp, eventType = 'web.webinteraction.linkClicks', 'minutes')
           OVER (PARTITION BY _{tenant_id}.user_id ORDER BY timestamp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
           AS "minutes_since_linkClick",
        row_number() OVER (PARTITION BY _{tenant_id}.user_id ORDER BY randn()) AS random_row_number_for_user
    FROM {table_name}
    SNAPSHOT BETWEEN $from_snapshot_id AND $to_snapshot_id
)
WHERE
    @my_table_exists = 't' AND
    ((subscriptionOccurred = 1 AND eventType = 'web.formFilledOut') OR (subscriptionOccurred = 0 AND random_row_number_for_user = 1))
ORDER BY timestamp;

EXCEPTION
  WHEN OTHER THEN
    SELECT 'ERROR';

END $$;
"""

Schließlich speichert der folgende Code die Abfragevorlage in Data Distiller:

template_res = dd.createQueryTemplate({
  "sql": query_training_set_template,
  "queryParameters": {},
  "name": "Template for propensity training data"
})
template_id = template_res["id"]

print(f"Template for propensity training data created as ID {template_id}")

Beispielausgabe

Template for propensity training data created as ID f3d1ec6b-40c2-4d13-93b6-734c1b3c7235

Nachdem die Vorlage gespeichert wurde, können Sie die Abfrage jederzeit ausführen, indem Sie auf die Vorlagen-ID verweisen und den Bereich der Momentaufnahme-IDs angeben, die in die Abfrage aufgenommen werden sollen. Die folgende Abfrage ruft die Momentaufnahmen des ursprünglichen Erlebnisereignis-Datensatzes ab:

query_snapshots = f"""
SELECT snapshot_id
FROM (
    SELECT history_meta('{table_name}')
)
WHERE is_current = true OR snapshot_generation = 0
ORDER BY snapshot_generation ASC
"""

df_snapshots = dd_cursor.query(query_snapshots, output="dataframe")

Der folgende Code veranschaulicht die Ausführung der Abfragevorlage unter Verwendung der ersten und letzten Momentaufnahmen zur Abfrage des gesamten Datensatzes:

snapshot_start_id = str(df_snapshots["snapshot_id"].iloc[0])
snapshot_end_id = str(df_snapshots["snapshot_id"].iloc[1])

query_final_res = qs.postQueries(
    name=f"[CMLE][Week2] Query to generate training data created by {username}",
    templateId=template_id,
    queryParameters={
        "from_snapshot_id": snapshot_start_id,
        "to_snapshot_id": snapshot_end_id,
    },
    dbname=f"{cat_conn.sandbox}:all"
)
query_final_id = query_final_res["id"]
print(f"Query started successfully and got assigned ID {query_final_id} - it will take some time to execute")

Beispielausgabe

Query started successfully and got assigned ID c6ea5009-1315-4839-b072-089ae01e74fd - it will take some time to execute

Sie können die folgende Funktion definieren, um den Status der Abfrage regelmäßig zu überprüfen:

def wait_for_query_completion(query_id):
    while True:
        query_info = qs.getQuery(query_id)
        query_state = query_info["state"]
        if query_state in ["SUCCESS", "FAILED"]:
            break
        print("Query is still in progress, sleeping…")
        time.sleep(60)

    duration_secs = query_info["elapsedTime"] / 1000
    if query_state == "SUCCESS":
        print(f"Query completed successfully in {duration_secs} seconds")
    else:
        print(f"Query failed with the following errors:", file=sys.stderr)
        for error in query_info["errors"]:
            print(f"Error code {error['code']}: {error['message']}", file=sys.stderr)

wait_for_query_completion(query_final_id)

Beispielausgabe

Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query is still in progress, sleeping…
Query completed successfully in 473.8 seconds

Nächste Schritte:

Durch das Lesen dieses Dokuments haben Sie gelernt, wie Sie Daten in Adobe Experience Platform in Funktionen oder Variablen umwandeln können, die von einem maschinellen Lernmodell genutzt werden können. Der nächste Schritt beim Erstellen von Funktions-Pipelines vom Experience Platform zum Einspeisen benutzerdefinierter Modelle in Ihrer maschinellen Lernumgebung besteht darin, Funktionsdatensätze zu exportieren.

recommendation-more-help
ccf2b369-4031-483f-af63-a93b5ae5e3fb