Merge branch 'machine-learning' into dev

src/documents/actions.py

@@ -239,9 +239,9 @@ def run_document_classifier_on_selected(modeladmin, request, queryset):
     n = queryset.count()
     if n:
         for obj in queryset:
-            clf.classify_document(obj, classify_correspondent=True, classify_tags=True, classify_type=True, replace_tags=True)
+            clf.classify_document(obj, classify_correspondent=True, classify_tags=True, classify_document_type=True, replace_tags=True)
             modeladmin.log_change(request, obj, str(obj))
-        modeladmin.message_user(request, "Successfully applied tags, correspondent and type to %(count)d %(items)s." % {
+        modeladmin.message_user(request, "Successfully applied tags, correspondent and document type to %(count)d %(items)s." % {
             "count": n, "items": model_ngettext(modeladmin.opts, n)
         }, messages.SUCCESS)
 

src/documents/apps.py

@@ -12,6 +12,7 @@ class DocumentsConfig(AppConfig):
         from .signals import document_consumption_finished
         from .signals.handlers import (
             classify_document,
+            add_inbox_tags,
             run_pre_consume_script,
             run_post_consume_script,
             cleanup_document_deletion,
@@ -21,6 +22,7 @@ class DocumentsConfig(AppConfig):
         document_consumption_started.connect(run_pre_consume_script)
 
         document_consumption_finished.connect(classify_document)
+        document_consumption_finished.connect(add_inbox_tags)
         document_consumption_finished.connect(set_log_entry)
         document_consumption_finished.connect(run_post_consume_script)
 

src/documents/classifier.py

@@ -2,12 +2,12 @@ import logging
 import os
 import pickle
 
+from sklearn.neural_network import MLPClassifier
+
 from documents.models import Correspondent, DocumentType, Tag, Document
 from paperless import settings
 
 from sklearn.feature_extraction.text import CountVectorizer
-from sklearn.multiclass import OneVsRestClassifier
-from sklearn.naive_bayes import MultinomialNB
 from sklearn.preprocessing import MultiLabelBinarizer, LabelBinarizer
 
 
@@ -29,11 +29,11 @@ class DocumentClassifier(object):
 
     tags_binarizer = None
     correspondent_binarizer = None
-    type_binarizer = None
+    document_type_binarizer = None
 
     tags_classifier = None
     correspondent_classifier = None
-    type_classifier = None
+    document_type_classifier = None
 
     @staticmethod
     def load_classifier():
@@ -48,11 +48,11 @@ class DocumentClassifier(object):
                 self.data_vectorizer = pickle.load(f)
                 self.tags_binarizer = pickle.load(f)
                 self.correspondent_binarizer = pickle.load(f)
-                self.type_binarizer = pickle.load(f)
+                self.document_type_binarizer = pickle.load(f)
 
                 self.tags_classifier = pickle.load(f)
                 self.correspondent_classifier = pickle.load(f)
-                self.type_classifier = pickle.load(f)
+                self.document_type_classifier = pickle.load(f)
             self.classifier_version = os.path.getmtime(settings.MODEL_FILE)
 
     def save_classifier(self):
@@ -61,33 +61,33 @@ class DocumentClassifier(object):
 
             pickle.dump(self.tags_binarizer, f)
             pickle.dump(self.correspondent_binarizer, f)
-            pickle.dump(self.type_binarizer, f)
+            pickle.dump(self.document_type_binarizer, f)
 
             pickle.dump(self.tags_classifier, f)
             pickle.dump(self.correspondent_classifier, f)
-            pickle.dump(self.type_classifier, f)
+            pickle.dump(self.document_type_classifier, f)
 
     def train(self):
         data = list()
         labels_tags = list()
         labels_correspondent = list()
-        labels_type = list()
+        labels_document_type = list()
 
         # Step 1: Extract and preprocess training data from the database.
         logging.getLogger(__name__).info("Gathering data from database...")
         for doc in Document.objects.exclude(tags__is_inbox_tag=True):
             data.append(preprocess_content(doc.content))
-            labels_type.append(doc.document_type.name if doc.document_type is not None and doc.document_type.automatic_classification else "-")
+            labels_document_type.append(doc.document_type.id if doc.document_type is not None and doc.document_type.automatic_classification else -1)
-            labels_correspondent.append(doc.correspondent.name if doc.correspondent is not None and doc.correspondent.automatic_classification else "-")
+            labels_correspondent.append(doc.correspondent.id if doc.correspondent is not None and doc.correspondent.automatic_classification else -1)
-            tags = [tag.name for tag in doc.tags.filter(automatic_classification=True)]
+            tags = [tag.id for tag in doc.tags.filter(automatic_classification=True)]
             labels_tags.append(tags)
 
         labels_tags_unique = set([tag for tags in labels_tags for tag in tags])
-        logging.getLogger(__name__).info("{} documents, {} tag(s) {}, {} correspondent(s) {}, {} type(s) {}.".format(len(data), len(labels_tags_unique), labels_tags_unique, len(set(labels_correspondent)), set(labels_correspondent), len(set(labels_type)), set(labels_type)))
+        logging.getLogger(__name__).info("{} documents, {} tag(s), {} correspondent(s), {} document type(s).".format(len(data), len(labels_tags_unique), len(set(labels_correspondent)), len(set(labels_document_type))))
 
         # Step 2: vectorize data
         logging.getLogger(__name__).info("Vectorizing data...")
-        self.data_vectorizer = CountVectorizer(analyzer='char', ngram_range=(2, 6), min_df=0.1)
+        self.data_vectorizer = CountVectorizer(analyzer='char', ngram_range=(3, 5), min_df=0.1)
         data_vectorized = self.data_vectorizer.fit_transform(data)
 
         self.tags_binarizer = MultiLabelBinarizer()
@@ -96,13 +96,13 @@ class DocumentClassifier(object):
         self.correspondent_binarizer = LabelBinarizer()
         labels_correspondent_vectorized = self.correspondent_binarizer.fit_transform(labels_correspondent)
 
-        self.type_binarizer = LabelBinarizer()
+        self.document_type_binarizer = LabelBinarizer()
-        labels_type_vectorized = self.type_binarizer.fit_transform(labels_type)
+        labels_document_type_vectorized = self.document_type_binarizer.fit_transform(labels_document_type)
 
         # Step 3: train the classifiers
         if len(self.tags_binarizer.classes_) > 0:
             logging.getLogger(__name__).info("Training tags classifier...")
-            self.tags_classifier = OneVsRestClassifier(MultinomialNB())
+            self.tags_classifier = MLPClassifier(verbose=True)
             self.tags_classifier.fit(data_vectorized, labels_tags_vectorized)
         else:
             self.tags_classifier = None
@@ -110,45 +110,58 @@ class DocumentClassifier(object):
 
         if len(self.correspondent_binarizer.classes_) > 0:
             logging.getLogger(__name__).info("Training correspondent classifier...")
-            self.correspondent_classifier = OneVsRestClassifier(MultinomialNB())
+            self.correspondent_classifier = MLPClassifier(verbose=True)
             self.correspondent_classifier.fit(data_vectorized, labels_correspondent_vectorized)
         else:
             self.correspondent_classifier = None
             logging.getLogger(__name__).info("There are no correspondents. Not training correspondent classifier.")
 
-        if len(self.type_binarizer.classes_) > 0:
+        if len(self.document_type_binarizer.classes_) > 0:
             logging.getLogger(__name__).info("Training document type classifier...")
-            self.type_classifier = OneVsRestClassifier(MultinomialNB())
+            self.document_type_classifier = MLPClassifier(verbose=True)
-            self.type_classifier.fit(data_vectorized, labels_type_vectorized)
+            self.document_type_classifier.fit(data_vectorized, labels_document_type_vectorized)
         else:
-            self.type_classifier = None
+            self.document_type_classifier = None
             logging.getLogger(__name__).info("There are no document types. Not training document type classifier.")
 
-    def classify_document(self, document, classify_correspondent=False, classify_type=False, classify_tags=False, replace_tags=False):
+    def classify_document(self, document, classify_correspondent=False, classify_document_type=False, classify_tags=False, replace_tags=False):
         X = self.data_vectorizer.transform([preprocess_content(document.content)])
 
         update_fields=()
 
         if classify_correspondent and self.correspondent_classifier is not None:
             y_correspondent = self.correspondent_classifier.predict(X)
-            correspondent = self.correspondent_binarizer.inverse_transform(y_correspondent)[0]
+            correspondent_id = self.correspondent_binarizer.inverse_transform(y_correspondent)[0]
-            print("Detected correspondent:", correspondent)
+            try:
-            document.correspondent = Correspondent.objects.filter(name=correspondent).first()
+                correspondent = Correspondent.objects.get(id=correspondent_id) if correspondent_id != -1 else None
-            update_fields = update_fields + ("correspondent",)
+                logging.getLogger(__name__).info("Detected correspondent: {}".format(correspondent.name if correspondent else "-"))
+                document.correspondent = correspondent
+                update_fields = update_fields + ("correspondent",)
+            except Correspondent.DoesNotExist:
+                logging.getLogger(__name__).warning("Detected correspondent with id {} does not exist anymore! Did you delete it?".format(correspondent_id))
 
-        if classify_type and self.type_classifier is not None:
+        if classify_document_type and self.document_type_classifier is not None:
-            y_type = self.type_classifier.predict(X)
+            y_type = self.document_type_classifier.predict(X)
-            type = self.type_binarizer.inverse_transform(y_type)[0]
+            type_id = self.document_type_binarizer.inverse_transform(y_type)[0]
-            print("Detected document type:", type)
+            try:
-            document.document_type = DocumentType.objects.filter(name=type).first()
+                document_type = DocumentType.objects.get(id=type_id) if type_id != -1 else None
-            update_fields = update_fields + ("document_type",)
+                logging.getLogger(__name__).info("Detected document type: {}".format(document_type.name if document_type else "-"))
+                document.document_type = document_type
+                update_fields = update_fields + ("document_type",)
+            except DocumentType.DoesNotExist:
+                logging.getLogger(__name__).warning("Detected document type with id {} does not exist anymore! Did you delete it?".format(type_id))
 
         if classify_tags and self.tags_classifier is not None:
             y_tags = self.tags_classifier.predict(X)
-            tags = self.tags_binarizer.inverse_transform(y_tags)[0]
+            tags_ids = self.tags_binarizer.inverse_transform(y_tags)[0]
-            print("Detected tags:", tags)
             if replace_tags:
                 document.tags.clear()
-            document.tags.add(*[Tag.objects.filter(name=t).first() for t in tags])
+            for tag_id in tags_ids:
+                try:
+                    tag = Tag.objects.get(id=tag_id)
+                    document.tags.add(tag)
+                    logging.getLogger(__name__).info("Detected tag: {}".format(tag.name))
+                except Tag.DoesNotExist:
+                    logging.getLogger(__name__).warning("Detected tag with id {} does not exist anymore! Did you delete it?".format(tag_id))
 
         document.save(update_fields=update_fields)

src/documents/management/commands/document_create_dataset.py

@@ -18,7 +18,7 @@ class Command(Renderable, BaseCommand):
         with open("dataset_tags.txt", "w") as f:
             for doc in Document.objects.exclude(tags__is_inbox_tag=True):
                 labels = []
-                for tag in doc.tags.all():
+                for tag in doc.tags.filter(automatic_classification=True):
                     labels.append(tag.name)
                 f.write(",".join(labels))
                 f.write(";")
@@ -27,14 +27,14 @@ class Command(Renderable, BaseCommand):
 
         with open("dataset_types.txt", "w") as f:
             for doc in Document.objects.exclude(tags__is_inbox_tag=True):
-                f.write(doc.document_type.name if doc.document_type is not None else "None")
+                f.write(doc.document_type.name if doc.document_type is not None and doc.document_type.automatic_classification else "-")
                 f.write(";")
                 f.write(preprocess_content(doc.content))
                 f.write("\n")
 
         with open("dataset_correspondents.txt", "w") as f:
             for doc in Document.objects.exclude(tags__is_inbox_tag=True):
-                f.write(doc.correspondent.name if doc.correspondent is not None else "None")
+                f.write(doc.correspondent.name if doc.correspondent is not None and doc.correspondent.automatic_classification else "-")
                 f.write(";")
                 f.write(preprocess_content(doc.content))
                 f.write("\n")

src/documents/management/commands/document_retagger.py

@@ -35,6 +35,10 @@ class Command(Renderable, BaseCommand):
             "-i", "--inbox-only",
             action="store_true"
         )
+        parser.add_argument(
+            "-r", "--replace-tags",
+            action="store_true"
+        )
 
     def handle(self, *args, **options):
 
@@ -52,7 +56,6 @@ class Command(Renderable, BaseCommand):
             logging.getLogger(__name__).fatal("Cannot classify documents, classifier model file was not found.")
             return
 
-
         for document in documents:
             logging.getLogger(__name__).info("Processing document {}".format(document.title))
-            clf.classify_document(document, classify_type=options['type'], classify_tags=options['tags'], classify_correspondent=options['correspondent'])
+            clf.classify_document(document, classify_document_type=options['type'], classify_tags=options['tags'], classify_correspondent=options['correspondent'], replace_tags=options['replace_tags'])

src/documents/signals/handlers.py

@@ -9,7 +9,7 @@ from django.contrib.contenttypes.models import ContentType
 from django.utils import timezone
 
 from documents.classifier import DocumentClassifier
-from ..models import Correspondent, Document, Tag, DocumentType
+from ..models import Document, Tag
 
 
 def logger(message, group):
@@ -23,11 +23,14 @@ def classify_document(sender, document=None, logging_group=None, **kwargs):
     global classifier
     try:
         classifier.reload()
-        classifier.classify_document(document, classify_correspondent=True, classify_tags=True, classify_type=True)
+        classifier.classify_document(document, classify_correspondent=True, classify_tags=True, classify_document_type=True)
     except FileNotFoundError:
         logging.getLogger(__name__).fatal("Cannot classify document, classifier model file was not found.")
 
 
+def add_inbox_tags(sender, document=None, logging_group=None, **kwargs):
+    inbox_tags = Tag.objects.filter(is_inbox_tag=True)
+    document.tags.add(*inbox_tags)
 
 
 def run_pre_consume_script(sender, filename, **kwargs):