project/backend/coordinator/controller/spacy_controller.py

@@ -5,17 +5,21 @@ import puremagic
 from werkzeug.utils import secure_filename
 from model.database import db
 import os
+import json
 import requests
 
 spacy_controller = Blueprint("spacy", __name__, url_prefix="/api/spacy")
 
 SPACY_TRAINING_URL = os.getenv("SPACY_TRAINING_URL", "http://spacy:5052/train")
-SPACY_URL = os.getenv("SPACY_URL", "http://spacy:5052")
+training_running_flag_path = os.path.join("spacy_training", "training_running.json")
 
 
 @spacy_controller.route("/train", methods=["POST"])
 def trigger_training():
     try:
+        with open(training_running_flag_path, "w") as f:
+            json.dump({"running": True}, f)
+
         response = requests.post(SPACY_TRAINING_URL, timeout=600)
         if response.ok:
             return jsonify({"message": "Training erfolgreich angestoßen."}), 200
@@ -105,29 +109,48 @@ def delete_file(id):
 
 
 @spacy_controller.route("/append-training-entry", methods=["POST"])
-def forward_training_entry():
+def append_training_entry():
     entry = request.get_json()
+
+    if not entry or "text" not in entry or "entities" not in entry:
+        return (
+            jsonify(
+                {"error": "Ungültiges Format – 'text' und 'entities' erforderlich."}
+            ),
+            400,
+        )
+
+    path = os.path.join("spacy_training", "annotation_data.json")
+
     try:
-        response = requests.post(f"{SPACY_URL}/append-training-entry", json=entry)
+        os.makedirs(os.path.dirname(path), exist_ok=True)
-        return jsonify(response.json()), response.status_code
+        if os.path.exists(path):
+            with open(path, "r", encoding="utf-8") as f:
+                data = json.load(f)
+        else:
+            data = []
+
+        if entry in data:
+            return jsonify({"message": "Eintrag existiert bereits."}), 200
+
+        data.append(entry)
+        with open(path, "w", encoding="utf-8") as f:
+            json.dump(data, f, indent=2, ensure_ascii=False)
+
+        return jsonify({"message": "Eintrag erfolgreich gespeichert."}), 200
     except Exception as e:
-        return jsonify({"error": str(e)}), 500
+        print(f"[ERROR] Fehler beim Schreiben: {e}")
-
+        return jsonify({"error": "Interner Fehler beim Schreiben."}), 500
-
-# globale Variable oben einfügen
-current_training_status = {"running": False}
-
-
-@spacy_controller.route("/training/status", methods=["POST"])
-def update_training_status():
-    data = request.get_json()
-    current_training_status["running"] = data.get("running", False)
-    running = current_training_status["running"]
-    print(f"[INFO] Trainingsstatus aktualisiert: running = {running}")
-
-    return jsonify({"status": "success", "running": current_training_status["running"]})
 
 
 @spacy_controller.route("/train-status", methods=["GET"])
 def training_status():
-    return jsonify(current_training_status), 200
+    try:
+        if os.path.exists(training_running_flag_path):
+            with open(training_running_flag_path, "r") as f:
+                status = json.load(f)
+            return jsonify(status), 200
+        else:
+            return jsonify({"running": False}), 200
+    except Exception as e:
+        return jsonify({"error": "Fehler beim Statuscheck", "details": str(e)}), 500

project/backend/exxetaGPT-service/Dockerfile

@@ -0,0 +1,8 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+COPY . /app
+RUN pip install --no-cache-dir -r requirements.txt
+ENV PYTHONUNBUFFERED=1
+
+CMD ["python", "extractExxeta.py"]

project/backend/spacy-service/app.py

@@ -14,7 +14,7 @@ training_status = {"running": False}
 app = Flask(__name__)
 CORS(app)
 
-COORDINATOR_URL = os.getenv("COORDINATOR_URL", "http://coordinator:5000")
+
 VALIDATE_SERVICE_URL = os.getenv(
     "VALIDATE_SERVICE_URL", "http://localhost:5054/validate"
 )
@@ -88,11 +88,31 @@ def append_training_entry():
 @app.route("/train", methods=["POST"])
 def trigger_training():
     from threading import Thread
+    import subprocess
+    import shutil
+
+    def run_training():
+        training_status["running"] = True
+        try:
+            if os.path.exists("output/model-last"):
+                shutil.copytree(
+                    "output/model-last", "output/model-backup", dirs_exist_ok=True
+                )
+            subprocess.run(["python", "spacy_training/ner_trainer.py"], check=True)
+            load_model()
+        except Exception as e:
+            print("Training failed:", e)
+        training_status["running"] = False
 
     Thread(target=run_training).start()
     return jsonify({"message": "Training gestartet"}), 200
 
 
+@app.route("/train-status", methods=["GET"])
+def get_training_status():
+    return jsonify(training_status), 200
+
+
 @app.route("/reload-model", methods=["POST"])
 def reload_model():
     try:
@@ -107,32 +127,16 @@ def reload_model():
 
 def run_training():
     training_status["running"] = True
-    notify_coordinator(True)
-
     try:
         if os.path.exists("output/model-last"):
             shutil.copytree(
                 "output/model-last", "output/model-backup", dirs_exist_ok=True
             )
         subprocess.run(["python", "spacy_training/ner_trainer.py"], check=True)
-        load_model()
+        load_model()  # ⬅ Modell nach dem Training direkt neu laden
     except Exception as e:
         print("Training failed:", e)
     training_status["running"] = False
-    notify_coordinator(False)
-
-
-def notify_coordinator(running: bool):
-    try:
-        response = requests.post(
-            f"{COORDINATOR_URL}/api/spacy/training/status", json={"running": running}
-        )
-        print(
-            f"[SPACY] Coordinator: running = {running}, Status = {response.status_code}"
-        )
-
-    except Exception as e:
-        print(f"[SPACY] Fehler beim Senden des Trainingsstatus: {e}")
 
 
 if __name__ == "__main__":

project/backend/spacy-service/spacy_training/annotation_data.json

@@ -1638,5 +1638,185 @@
         "RENDITE"
       ]
     ]
+  },
+  {
+    "text": "Die Gesamtrendite beträgt 7,2 %.",
+    "entities": [
+      [
+        1,
+        5,
+        "NEUEKENNZAHL"
+      ]
+    ]
+  },
+  {
+    "text": "fhfhfh56",
+    "entities": [
+      [
+        6,
+        8,
+        "TEST545"
+      ]
+    ]
+  },
+  {
+    "text": "fhfhfh56",
+    "entities": [
+      [
+        6,
+        8,
+        "TEST345"
+      ]
+    ]
+  },
+  {
+    "text": "sdgds45",
+    "entities": [
+      [
+        6,
+        7,
+        "TEST243"
+      ]
+    ]
+  },
+  {
+    "text": "4t4r3",
+    "entities": [
+      [
+        4,
+        5,
+        "TEST243"
+      ]
+    ]
+  },
+  {
+    "text": "sdgds45",
+    "entities": [
+      [
+        6,
+        7,
+        "DGTDDTFHZ"
+      ]
+    ]
+  },
+  {
+    "text": "gjufzj45",
+    "entities": [
+      [
+        7,
+        8,
+        "DGTDDTFHZ"
+      ]
+    ]
+  },
+  {
+    "text": "irr beträgt 43",
+    "entities": [
+      [
+        12,
+        14,
+        "TEST3243"
+      ]
+    ]
+  },
+  {
+    "text": "irr beträgt 43",
+    "entities": [
+      [
+        12,
+        14,
+        "IRR"
+      ]
+    ]
+  },
+  {
+    "text": "Rendite besträgt 5 %",
+    "entities": [
+      [
+        17,
+        20,
+        "RENDITE"
+      ]
+    ]
+  },
+  {
+    "text": "RenditeX besträgt 5 %",
+    "entities": [
+      [
+        18,
+        21,
+        "RENDITE_X"
+      ]
+    ]
+  },
+  {
+    "text": "gtg3ahz8",
+    "entities": [
+      [
+        7,
+        8,
+        "ERTRETT"
+      ]
+    ]
+  },
+  {
+    "text": "wffwee 45",
+    "entities": [
+      [
+        7,
+        9,
+        "TEST45"
+      ]
+    ]
+  },
+  {
+    "text": "efwwef 45",
+    "entities": [
+      [
+        7,
+        9,
+        "TEST12"
+      ]
+    ]
+  },
+  {
+    "text": "wfwefwe34",
+    "entities": [
+      [
+        7,
+        9,
+        "TEST232"
+      ]
+    ]
+  },
+  {
+    "text": "fwefbmj34",
+    "entities": [
+      [
+        7,
+        9,
+        "TEST223"
+      ]
+    ]
+  },
+  {
+    "text": "asdas45",
+    "entities": [
+      [
+        5,
+        7,
+        "TEST122"
+      ]
+    ]
+  },
+  {
+    "text": "ewefw4",
+    "entities": [
+      [
+        5,
+        6,
+        "TEST3434"
+      ]
+    ]
   }
 ]

project/backend/spacy-service/spacy_training/base_config.cfg

@@ -0,0 +1,85 @@
+# This is an auto-generated partial config. To use it with 'spacy train'
+# you can run spacy init fill-config to auto-fill all default settings:
+# python -m spacy init fill-config ./base_config.cfg ./config.cfg
+[paths]
+train = ./data/train.spacy
+dev = ./data/train.spacy
+vectors = null
+[system]
+gpu_allocator = null
+
+[nlp]
+lang = "de"
+pipeline = ["tok2vec","ner"]
+batch_size = 1000
+
+[components]
+
+[components.tok2vec]
+factory = "tok2vec"
+
+[components.tok2vec.model]
+@architectures = "spacy.Tok2Vec.v2"
+
+[components.tok2vec.model.embed]
+@architectures = "spacy.MultiHashEmbed.v2"
+width = ${components.tok2vec.model.encode.width}
+attrs = ["NORM", "PREFIX", "SUFFIX", "SHAPE"]
+rows = [5000, 1000, 2500, 2500]
+include_static_vectors = false
+
+[components.tok2vec.model.encode]
+@architectures = "spacy.MaxoutWindowEncoder.v2"
+width = 96
+depth = 4
+window_size = 1
+maxout_pieces = 3
+
+[components.ner]
+factory = "ner"
+
+[components.ner.model]
+@architectures = "spacy.TransitionBasedParser.v2"
+state_type = "ner"
+extra_state_tokens = false
+hidden_width = 64
+maxout_pieces = 2
+use_upper = true
+nO = null
+
+[components.ner.model.tok2vec]
+@architectures = "spacy.Tok2VecListener.v1"
+width = ${components.tok2vec.model.encode.width}
+
+[corpora]
+
+[corpora.train]
+@readers = "spacy.Corpus.v1"
+path = ${paths.train}
+max_length = 0
+
+[corpora.dev]
+@readers = "spacy.Corpus.v1"
+path = ${paths.dev}
+max_length = 0
+
+[training]
+dev_corpus = "corpora.dev"
+train_corpus = "corpora.train"
+
+[training.optimizer]
+@optimizers = "Adam.v1"
+
+[training.batcher]
+@batchers = "spacy.batch_by_words.v1"
+discard_oversize = false
+tolerance = 0.2
+
+[training.batcher.size]
+@schedules = "compounding.v1"
+start = 100
+stop = 1000
+compound = 1.001
+
+[initialize]
+vectors = ${paths.vectors}

project/backend/spacy-service/spacy_training/config.cfg

@@ -0,0 +1,145 @@
+[paths]
+train = "./data/train.spacy"
+dev = "./data/train.spacy"
+vectors = null
+init_tok2vec = null
+
+[system]
+gpu_allocator = null
+seed = 0
+
+[nlp]
+lang = "de"
+pipeline = ["tok2vec","ner"]
+batch_size = 1000
+disabled = []
+before_creation = null
+after_creation = null
+after_pipeline_creation = null
+tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
+vectors = {"@vectors":"spacy.Vectors.v1"}
+
+[components]
+
+[components.ner]
+factory = "ner"
+incorrect_spans_key = null
+moves = null
+scorer = {"@scorers":"spacy.ner_scorer.v1"}
+update_with_oracle_cut_size = 100
+
+[components.ner.model]
+@architectures = "spacy.TransitionBasedParser.v2"
+state_type = "ner"
+extra_state_tokens = false
+hidden_width = 64
+maxout_pieces = 2
+use_upper = true
+nO = null
+
+[components.ner.model.tok2vec]
+@architectures = "spacy.Tok2VecListener.v1"
+width = ${components.tok2vec.model.encode.width}
+upstream = "*"
+
+[components.tok2vec]
+factory = "tok2vec"
+
+[components.tok2vec.model]
+@architectures = "spacy.Tok2Vec.v2"
+
+[components.tok2vec.model.embed]
+@architectures = "spacy.MultiHashEmbed.v2"
+width = ${components.tok2vec.model.encode.width}
+attrs = ["NORM","PREFIX","SUFFIX","SHAPE"]
+rows = [5000,1000,2500,2500]
+include_static_vectors = false
+
+[components.tok2vec.model.encode]
+@architectures = "spacy.MaxoutWindowEncoder.v2"
+width = 96
+depth = 4
+window_size = 1
+maxout_pieces = 3
+
+[corpora]
+
+[corpora.dev]
+@readers = "spacy.Corpus.v1"
+path = ${paths.dev}
+max_length = 0
+gold_preproc = false
+limit = 0
+augmenter = null
+
+[corpora.train]
+@readers = "spacy.Corpus.v1"
+path = ${paths.train}
+max_length = 0
+gold_preproc = false
+limit = 0
+augmenter = null
+
+[training]
+dev_corpus = "corpora.dev"
+train_corpus = "corpora.train"
+seed = ${system.seed}
+gpu_allocator = ${system.gpu_allocator}
+dropout = 0.1
+accumulate_gradient = 1
+patience = 1600
+max_epochs = 0
+max_steps = 20000
+eval_frequency = 200
+frozen_components = []
+annotating_components = []
+before_to_disk = null
+before_update = null
+
+[training.batcher]
+@batchers = "spacy.batch_by_words.v1"
+discard_oversize = false
+tolerance = 0.2
+get_length = null
+
+[training.batcher.size]
+@schedules = "compounding.v1"
+start = 100
+stop = 1000
+compound = 1.001
+t = 0.0
+
+[training.logger]
+@loggers = "spacy.ConsoleLogger.v1"
+progress_bar = false
+
+[training.optimizer]
+@optimizers = "Adam.v1"
+beta1 = 0.9
+beta2 = 0.999
+L2_is_weight_decay = true
+L2 = 0.01
+grad_clip = 1.0
+use_averages = false
+eps = 0.00000001
+learn_rate = 0.001
+
+[training.score_weights]
+ents_f = 1.0
+ents_p = 0.0
+ents_r = 0.0
+ents_per_type = null
+
+[pretraining]
+
+[initialize]
+vectors = ${paths.vectors}
+init_tok2vec = ${paths.init_tok2vec}
+vocab_data = null
+lookups = null
+before_init = null
+after_init = null
+
+[initialize.components]
+
+[initialize.tokenizer]

project/backend/spacy-service/spacy_training/data_to_json.py

@@ -0,0 +1,18 @@
+import os
+import json
+from training_data import TRAINING_DATA
+
+# Setze hier den Pfad zu annotation_data.json
+OUTFILE = os.path.join(os.path.dirname(__file__), "annotation_data.json")
+
+json_list = []
+for text, annot in TRAINING_DATA:
+    entities = []
+    for start, end, label in annot["entities"]:
+        entities.append([start, end, label])
+    json_list.append({"text": text, "entities": entities})
+
+with open(OUTFILE, "w", encoding="utf8") as f:
+    json.dump(json_list, f, ensure_ascii=False, indent=2)
+
+print("Alle Trainingsdaten wurden erfolgreich nach annotation_data.json migriert!")

project/backend/spacy-service/spacy_training/output/model-last/meta.json

@@ -19,13 +19,16 @@
   "labels":{
     "ner":[
       "AUSSCH\u00dcTTUNGSRENDITE",
+      "IRR",
       "KENNZAHL",
       "LAUFZEIT",
       "L\u00c4NDERALLOKATION",
       "MANAGMENTGEB\u00dcHREN",
       "RENDITE",
+      "RENDITE_X",
       "RISIKOPROFIL",
       "SEKTORENALLOKATION",
+      "TEST3243",
       "ZIELAUSSCH\u00dcTTUNG",
       "ZIELRENDITE"
     ]

project/backend/spacy-service/spacy_training/output/model-last/ner/moves

@@ -1 +1 @@
-‚¥movesÚˆ{"0":{},"1":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10},"2":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10},"3":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10},"4":{"":1,"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10},"5":{"":1}}£cfg<66>§neg_keyÀ
+‚¥movesÚ,{"0":{},"1":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10,"TEST3243":-11,"IRR":-12,"RENDITE_X":-13},"2":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10,"TEST3243":-11,"IRR":-12,"RENDITE_X":-13},"3":{"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10,"TEST3243":-11,"IRR":-12,"RENDITE_X":-13},"4":{"":1,"KENNZAHL":-1,"RISIKOPROFIL":-2,"AUSSCH\u00dcTTUNGSRENDITE":-3,"LAUFZEIT":-4,"RENDITE":-5,"L\u00c4NDERALLOKATION":-6,"ZIELRENDITE":-7,"ZIELAUSSCH\u00dcTTUNG":-8,"MANAGMENTGEB\u00dcHREN":-9,"SEKTORENALLOKATION":-10,"TEST3243":-11,"IRR":-12,"RENDITE_X":-13},"5":{"":1}}£cfg<66>§neg_keyÀ

project/backend/spacy-service/spacy_training/output/model-last/vocab/strings.json

@@ -273,8 +273,11 @@
   "4,91",
   "40",
   "400",
+  "43",
   "45",
   "491",
+  "4r3",
+  "4t4r3",
   "5",
   "5%+",
   "5,0",
@@ -305,6 +308,7 @@
   "67",
   "7",
   "7,1",
+  "7,2",
   "7,5",
   "7,5%+",
   "7,50",
@@ -1065,6 +1069,7 @@
   "R.I.P.",
   "RE",
   "RENDITE",
+  "RENDITE_X",
   "REV",
   "REWE",
   "RISIKOPROFIL",
@@ -1079,8 +1084,10 @@
   "Redaktion",
   "Region",
   "Regionen",
+  "Rendite",
   "Rendite-",
   "Rendite-Risiko-Profil",
+  "RenditeX",
   "Renovierungen",
   "Rents",
   "Residential",
@@ -1160,6 +1167,7 @@
   "T",
   "T.",
   "TED",
+  "TEST3243",
   "Tag",
   "Target",
   "Target-IRR",
@@ -1300,6 +1308,7 @@
   "Xxxxx-Xxxxx-Xxxxx",
   "Xxxxx-xxx",
   "Xxxxx-xxxx",
+  "XxxxxX",
   "Xxxxx\u0308xx",
   "Xxxxx\u0308xxx-Xxxxx",
   "Xxxxx\u0308xxxx",
@@ -1394,7 +1403,6 @@
   "across",
   "act",
   "active",
-  "adasd23",
   "add",
   "adv",
   "adv.",
@@ -1551,9 +1559,11 @@
   "berlin",
   "bestandsentwicklung",
   "bestandsentwicklungen",
+  "bestr\u00e4gt",
   "betr",
   "betr.",
   "betreute",
+  "betr\u00e4gt",
   "bev\u00f6lkerungsprognose",
   "beziehungsweise",
   "bez\u00fcglich",
@@ -1684,7 +1694,6 @@
   "d.h",
   "d.h.",
   "d.x",
-  "d23",
   "dX",
   "dXxx.\u20ac",
   "d_d",
@@ -1765,6 +1774,7 @@
   "durchschnittlich",
   "du\u2019s",
   "dv.",
+  "dxdxd",
   "dy",
   "d\u00e4nemark",
   "d\u2019",
@@ -1929,6 +1939,7 @@
   "festgelegt",
   "festgelegter",
   "ff",
+  "fhfhfh56",
   "fierce",
   "fil",
   "financially",
@@ -2029,6 +2040,7 @@
   "ght",
   "gic",
   "gie",
+  "gjufzj45",
   "gl.",
   "global",
   "globale",
@@ -2050,6 +2062,7 @@
   "h.",
   "h.c",
   "h.c.",
+  "h56",
   "haltedauer",
   "halten",
   "halten-strategie",
@@ -2249,6 +2262,7 @@
   "ize",
   "j",
   "j.",
+  "j45",
   "ja",
   "jahr",
   "jahre",
@@ -2729,8 +2743,10 @@
   "relationships",
   "remains",
   "ren",
+  "rendite",
   "rendite-",
   "rendite-risiko-profil",
+  "renditex",
   "renegotiation",
   "renovierungen",
   "rent",
@@ -2787,6 +2803,7 @@
   "s.o",
   "s.o.",
   "s.w",
+  "s45",
   "sa",
   "sa.",
   "sale",
@@ -2797,6 +2814,7 @@
   "scs",
   "scsp",
   "sd.",
+  "sdgds45",
   "sector",
   "sectors",
   "sed",
@@ -2902,6 +2920,7 @@
   "tc.",
   "td.",
   "te-",
+  "teX",
   "ted",
   "tee",
   "teflimmobilfe)-",
@@ -3227,6 +3246,7 @@
   "\u00e4",
   "\u00e4.",
   "\u00e4gl",
+  "\u00e4gt",
   "\u00e4r.",
   "\u00e4rzteh\u00e4user",
   "\u00e4rzteh\u00e4usern",

project/backend/spacy-service/spacy_training/test.json

@@ -0,0 +1,9 @@
+{
+  "id": "TEST",
+  "extracted_text_per_page": [
+    {
+      "page": 1,
+      "text": "Die Gesamtrendite beträgt 7,2 %."
+    }
+  ]
+}

project/backend/spacy-service/spacy_training/training_data.py

@@ -0,0 +1,563 @@
+TRAINING_DATA = [
+    (
+        "Core",
+        {"entities": [[0, 4, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core+",
+        {"entities": [[0, 5, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core/Core+",
+        {"entities": [[0, 10, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Value Add",
+        {"entities": [[0, 9, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core/Value Add",
+        {"entities": [[0, 14, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core+/Value Add",
+        {"entities": [[0, 15, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core/Core+/Value Add",
+        {"entities": [[0, 20, "RISIKOPROFIL"]]},
+    ),
+    (
+        "The RE portfolio of the fund is a good illustration of Fond expertise in European core/core+ investments .",
+        {"entities": [[82, 92, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Risk level: Core/Core+",
+        {"entities": [[12, 22, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Different risk profile (core, core+, value-added)",
+        {"entities": [[24, 48, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core/Core+ with OpCo premium",
+        {"entities": [[0, 10, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core /Core+ Assets, well-established = Key Gateway Cities in Europe le.g. hotels in the market with minor asset London, Paris, Amsterdam, Berlin] management initiatives",
+        {"entities": [[0, 11, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Risikoprofil: Core, Core +",
+        {"entities": [[14, 26, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Name des Fonds Name des Investmentmanagers Allgemeine Informationen Name des Ansprechpartners Telefonnummer des Ansprechpartners E-Mail des Ansprechpartners Art des Anlagevehikels Struktur des Anlagevehikels Sitz des Anlagevehikels Struktur des Antagevehikels vom Manager festgelegter Stil Rechtsform Jahr des ersten Closings Laufzeit Geplantes Jahr der Auflösung Ziel-Netto-IRR / Gesamtrendite* Zielvolumen des Anlagevehikels Ziel-LTY ‚Aktueller LTV Ziirraiaein Maximaler LTV Zielregionfen)/Jand Zielsektoren Zielanlagestrategie INREV Fonds Offen Deutschland Core, Core + Offener Immobilien-Spezialfonds 2022 10 - 12 Jahre 2032 - 2034 7,50%+ 250 Mio. € 20% 0% 20% Führende Metropolregionen Deutschlands und ausgewählte Standorte >50T Einw. Wohnimmobilien Wertstabile Wohnimmobilien (mit Bestandsentwicklungen)",
+        {"entities": [[560, 572, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core/Core+ strategy, with tactical exposure to development projects aiming at enhancing the quality of the portfolio over time",
+        {"entities": [[0, 10, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Strategie - Übersicht Risikoprofil Core+ Halten-Strategie Kaufen — Halten (langfristig) — Exit 1. Nachvermietungsstrategie Anlagestrategien 2. Standortaufwertungsstrategie 3. Strategie der Aufwertung der Immobilien Niederlande (max. 35 %) Länderallokation Frankreich (max. 35 %) (in % vom Zielvolumen) Skandinavien (Schweden, Dänemark) (max. 35 %) Deutschland (<= 10 %)",
+        {"entities": [[35, 40, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core and Core+",
+        {"entities": [[0, 14, "RISIKOPROFIL"]]},
+    ),
+    (
+        "core, core+, value-added",
+        {"entities": [[0, 24, "RISIKOPROFIL"]]},
+    ),
+     (
+         "Manage to Core: max 20%",
+         {"entities": [[10, 14, "RISIKOPROFIL"]]},
+     ),
+     (
+         "Benefits of the core/ core+ segment",
+         {"entities": [[16, 27, "RISIKOPROFIL"]]},
+     ),
+     (
+         "Drawbacks of the core/ core+ segment",
+         {"entities": [[17, 28, "RISIKOPROFIL"]]},
+     ),
+     (
+         "Why a Core / Core + investment program?",
+         {"entities": [[6, 19, "RISIKOPROFIL"]]},
+     ),
+     (
+         "Different risk profile (core, core+, value-added)",
+         {"entities": [[24, 48, "RISIKOPROFIL"]]},
+     ),
+     (
+        "INK MGallery Hotel Area: Amsterdam Core Tenant: Closed in 2018",
+        {"entities": [[35, 39, "RISIKOPROFIL"]]},
+     ),
+     (
+         "A strategy targeting high quality Core and Core+ buildings, with defined SRI objectives, in order to extract value through an active asset management.",
+         {"entities": [[34, 48, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Navigate the diversity of the Core/Core+ investment opportunities in European Prime Cities",
+        {"entities": [[30, 40, "RISIKOPROFIL"]]},
+    ),
+    (
+        "GEDis an open-ended Lux-based fund providing an attractive core/core+ real estate exposure, leveraging GRRE expertise in European RE markets. It offers diversification in terms of pan-European geographies and sectors: Offices, Retail and Hotels.",
+        {"entities": [[59, 69, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core assets leave less room for active asset management value creation",
+        {"entities": [[0, 4, "RISIKOPROFIL"]]},
+    ),
+    (
+        "capital preservation is defined here as a characteristic of core/core+ investments. There is no guarantee of capital.",
+        {"entities": [[60, 70, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Country / city BELGIUM Brussels BELGIUM Brussels SPAIN Madrid FRANCE Levallois FRANCE Paris 14 BELGIUM Brussels NETHERLANDS Rotterdam NETHERLANDS Rotterdam Sector Offices Offices Offices Offices Offices Offices Offices Logistics Risk Core",
+        {"entities": [[234, 238, "RISIKOPROFIL"]]},
+    ),
+    (
+        "GERD(a balanced pan-European open ended retail fund — under the form of a French collective undertaking for Real Estate investments “OPCI”) is the flagship ofQin France and combines RE and listed assets (respective targets of 60% and 40%) with max. 40% leverage. The RE portfolio of the fund is a good illustration Of expertise in European core/core+ investments.",
+        {"entities": [[340, 350, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Prime office assets in Prime markets are very pricey unless rent reversion is real. Risk premium remains attractive on a leveraged basis. Manage to core or build to core can make sense as a LT investor in main cities. Residential is also attractive",
+        {"entities": [[148, 152, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Paris region is a deep and liquid market. Rents have some potential to improve. Considering current low yield and fierce competition, office right outside CBD for Core + assets can be considered. Manage to core strategies could make sense.",
+        {"entities": [[163, 169, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Lisbon is a small market but it experienced a rapid economic recovery in recent years and is interesting for Core Offices, quality Retail assetor Hotel walls with top operators. Limited liquidity of this market means investment must be small",
+        {"entities": [[109, 113, "RISIKOPROFIL"]]},
+    ),
+    (
+        "4,0 %",
+        {"entities": [[0, 5, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Prognostizierte jährliche Ausschüttung von 4,0%",
+        {"entities": [[44, 48, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "20% über einer @ Ausschüttungsrendite von 4,0%",
+        {"entities": [[44, 48, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Prognostizierte Ausschüttungsrandite* Mindestanlage Mitgliedschaft Im Anlagesusschuss Ankaufs- / Verkaufs- / Verkaufs(Teflimmobilfe)- / Baumanagementgebahr (inkl. USt.) Parformanceabhängige Vergütung Einmalige Strukturierungsgebühr Laufzeit / Investtionszeltraum Ausschüttungsintervalle Deutsche Metropolregianen und umliegende Regionen mit Städten >50T Einwohner Artikel 8 Wohnimmobilien Deutschland ‚Aktive Bestandsentwicklung Offener Spezial-AlF mit festen Anlagebedingungen rd. 200 Mio. € / max. 20% rd. 250 Mio. € 7,5 % (nach Kosten & Gebühren, vor Steuern) 8 4,0 % {nach Kosten & Gebühren, var Steuern}",
+        {"entities": [[570, 575, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "5,00-5,25 % Ausschüttungsrendite",
+        {"entities": [[0, 11, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Zielrendite 5,00-5,25 % Ausschüttungsrendite",
+        {"entities": [[12, 23, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 4,9% 5,3%",
+        {"entities": [[21, 25, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 4,9% 5,3%",
+        {"entities": [[26, 30, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschittungsrendite 3,8% 5,7%",
+        {"entities": [[20, 24, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschittungsrendite 3,8% 5,7%",
+        {"entities": [[25, 29, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 4,5% 4,6%",
+        {"entities": [[21, 25, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 4,5% 4,6%",
+        {"entities": [[26, 30, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 5,0% 4,7%",
+        {"entities": [[26, 30, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite 5,0% 4,7%",
+        {"entities": [[21, 25, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite “eons a Nuremberg aha 5,0 % 4,8 %",
+        {"entities": [[43, 48, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Auschüttungsrendite “eons a Nuremberg aha 5,0 % 4,8 %",
+        {"entities": [[49, 54, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "3-4% dividend yield",
+        {"entities": [[0, 4, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Zielmärkte Klassifizierung SFDR Invastitionsfokus Rendite- / Risikoprofil Rechtsform Eigenkapital /FK Quote Investftionsvolumen Prognostizierte Gesamtrendite {IRR)* Prognostizierte Ausschüttungsrandite* Mindestanlage Mitgliedschaft Im Anlagesusschuss Ankaufs- / Verkaufs- / Verkaufs(Teflimmobilfe)- / Baumanagementgebahr (inkl. USt.) Parformanceabhängige Vergütung Einmalige Strukturierungsgebühr Deutsche Metropolregianen und umliegende Regionen mit Städten >50T Einwohner Artikel 8 Wohnimmobilien Deutschland ‚Aktive Bestandsentwicklung Offener Spezial-AlF mit festen Anlagebedingungen rd. 200 Mio. € / max. 20% rd. 250 Mio. € 7,5 % (nach Kosten & Gebühren, vor Steuern) 8 4,0 % {nach Kosten & Gebühren, var Steuern} 5Mio.€ Ab 10 Mio. € 1,40 % / 0,80 % /2,12% / 4,91 % Laufzeit / Investtionszeltraum Ausschüttungsintervalle 20 % über einer @ Ausschüttungsrendite von 4,0 % 0,1% der bis zum 31.12.2023 erfolgten Kapitalzusagen (max. 200.000 &) 10 bis 12 Jahre / bis zu 24 Monate angestrebt Mindestens jährlich",
+        {"entities": [[945, 960, "LAUFZEIT"]]},
+    ),
+    (
+        "Laufzeit / Investtionszeltraum,10 bis 12 Jahre / bis zu 24 Monate angestrebt Ausschüttungsintervalle,Mindestens jährlich",
+        {"entities": [[31, 46, "LAUFZEIT"]]},
+    ),
+    (
+        "10-12 Jahre Laufzeit bei einem LTV von bis zu 20%",
+        {"entities": [[0, 11, "LAUFZEIT"]]},
+    ),
+    (
+        "vom Manager festgelegter Stil Rechtsform Jahr des ersten Closings Laufzeit Geplantes Jahr der Auflösung Ziel-Netto-IRR / Gesamtrendite* Zielvolumen des Anlagevehikels Ziel-LTY‚Aktueller LTV Zielsektoren Zielanlagestrategie Fonds Offen Deutschland Core, Core + Offener Immobilien-Spezialfonds 2022 10 - 12 Jahre",
+        {"entities": [[297, 310, "LAUFZEIT"], [247, 259, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Allgemeine Annahmen Ankaufsphase Haltedauer Zielobjektgröße Finanzierung Investitions-annahmen Zielrendite 24 Monate Investmentzeitraum 10 Jahre (+) EUR 20-75 Mio. Keine externe Finanzierung zum Auftakt (ausschließlich Darlehen der Anteilseigner). Die Finanzierung wird nach der Ankaufsphase und Stabilisierung der Zinssätze neu geprüft. Angestrebter LTV zwischen 25-40 % Investitionen für Renovierungen und ESG- Verbesserungen werden für jedes Objekt einzeln festgelegt. 5,00-5,25 % Ausschüttungsrendites",
+        {"entities": [[136, 148, "LAUFZEIT"], [472, 483, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Zielrendite 5,00-5,25 % Ausschüttungsrendite 1) Ankauf von Objekten an Tag eins mit 100% Eigenkapital. Die Strategie unterstellt die Aufnahme von Fremdkapital, sobald sich die Zins- und Finanzierungskonditionen nachhaltig stabilisieren. Strategie - Übersicht Risikoprofil Core+",
+        {"entities": [[12, 23, "AUSSCHÜTTUNGSRENDITE"], [272, 277, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Vehicle lifetime / investment period Open-ended fund",
+        {"entities": [[37, 52, "LAUFZEIT"]]},
+    ),
+    (
+        "Vehicle / domicile Alternative Investment Fund / Luxembourg (e.g. SCSp SICAV-RAIF) Investment strategy eturn pro Real Estate (PropCo + OpCo) Investing in upscale hotels with long-term management contracts in major European destinations Core/Core+ with OpCo premium Management Agreements solely with financially strong and experienced partners/ global brands Cash flow-oriented Cash-flow pattern Target equity /AuM € 400m equity / € 800m AuM (50% Loan-to-Value) Vehicle lifetime / investment period Open-ended fund",
+        {"entities": [[498, 513, "LAUFZEIT"], [236, 245, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Vehicle type (Lux-RAIF) (net of fees) IRR6.5% ACCOR Vehicle structure Open-ended Targetvehiclesize € 400m (equity) Manager-defined Core/Core+ with | style OpCo Premium darge CLV. 50% Pt H | LTO N WORLDWIDE Year of first closing 2020 Target no. ofinvestors 1-5 Fund life (yrs} Open-ended Min-commitmentper —¢ 400m",
+        {"entities": [[131, 141, "RISIKOPROFIL"], [70, 80, "LAUFZEIT"]]},
+    ),
+    (
+        "Fund term: Open-ended",
+        {"entities": [[11, 21, "LAUFZEIT"]]},
+    ),
+    (
+        "Abdeckung der Risiko-Rendite-Bandbreite (Core, Core+, Value-Add)",
+        {"entities": [[41, 63, "RISIKOPROFIL"]]},
+    ),
+    (
+        "5,1% - 8,5% IRR!",
+        {"entities": [[0, 11, "RENDITE"]]},
+    ),
+    (
+        "Retailinvestitionsvolumen nach Ländern (2024) Vereinigtes Königreich, 26,4% Deutschland, 19,0% Andere, 19,7% Italien, 8,2% Irland, 3,3% N | Frankreich, Spanien, 8,1%",
+        {"entities": [[46, 75, "LÄNDERALLOKATION"], [76, 94, "LÄNDERALLOKATION"], [95, 108, "LÄNDERALLOKATION"], [109, 122, "LÄNDERALLOKATION"], [123, 135, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Erwartete IRR 5 (je nach Objekt- ‘A(E) 6.00% - 8,00%",
+        {"entities": [[39, 52, "RENDITE"]]},
+    ),
+    (
+        "Zielmarkts Deutsche Metropolregianen und umliegende Regionen mit Städten >50T Einwohner Klassifizierung SFDR Artikel 8 Invastitionsfokus Wohnimmobilien Deutschland Rendite- / Risikoprofil ‚Aktive Bestandsentwicklung Rechtsform Offener Spezial-AlF mit festen Anlagebedingungen Eigenkapital /FK Quote rd. 200 Mio. € / max. 20% Investftionsvolumen rd. 250 Mio. € Prognostiderte Gesamtrendite {IRR)* 7,5 % (nach Kosten & Gebühren, vor Steuern) Prognostizierte Ausschüttungsrandite* @ 4,0 % {nach Kosten & Gebühren, var Steuern} Mindestanlage 5Mio.€ Mitgliedschaft Im Anlagesusschuss Ab 10 Mio. € Ankaufs- / Verkaufs- / Verkaufs(Teflimmobilfe)- / Baumanagementgebahr (inkl. USt) 1,40 %/080%/212%/491% Parformanceabhängige Vergütung 20 % über einer ® Ausschüttungsrendite von 4,0% Einmalige Strukturierungsgebühr 0,1% der bis zum 31.12.2023 erfolgten Kapitalzusagen (max. 200.000 €) Laufzelt / Investtonszeltraum 10 bis 12 Jahre / bis zu 24 Monate angestrebt Ausschüttungsintervalle Mindestens jährlich",
+        {"entities": [[396, 401, "RENDITE"], [482, 487, "AUSSCHÜTTUNGSRENDITE"], [914, 929, "LAUFZEIT"]]},
+    ),
+    (
+        "= Prognostizierte jährliche Ausschüttung von @ 4,0%* = Prognostizierte Gesamtrendite (IRR) von 7,5%*",
+        {"entities": [[48, 52, "AUSSCHÜTTUNGSRENDITE"], [96, 100, "RENDITE"]]},
+    ),
+    (
+        "Prognose: 7,5%+ IRR auf Fondsebene",
+        {"entities": [[10, 14, "RENDITE"]]},
+    ),
+    (
+        "= Prognostizierte jährliche Ausschüttung* von 84,0% = Prognostizierte Gesamtrendite (IRR}* von 7,5%",
+        {"entities": [[96, 100, "RENDITE"], [49, 53, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "= Lagefokussierung: Metropolregionen Deutschlands = Finanzierung: max. 20% LTV = Risikoprofil: Core, Core +",
+        {"entities": [[95, 107, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Performance-Fee: 20% über einer @ Ausschüttungsrendite von 4,0%",
+        {"entities": [[61, 65, "AUSSCHÜTTUNGSRENDITE"]]},
+    ),
+    (
+        "Fondstyp Offener Spezial-AIF nach KAGB mit festen Anlagebedingungen ESG-Klassifizierung Fonds gemäß Artikel 8 EU-Offenlegungsverordnung KVG IntReal GmbH, Hamburg Anlagestrategie Aufbau eines Objektportfolios aus Ärztehäusern, die langfristig vermietet sind Ärztehäuser, Laborimmobilien, im Verbund mit Ärztehäusern auch ambulant Zielobjekte betreute Wohngemeinschaften; Mietanteil Medizin und medizinnahe Dienstleistungen/Handel > 65 % (Objektebene) WALT >5 Jahre bei Ankauf Objektbaujahre Ab 2000 Anlagegrenzen Einzelinvestment 8-30 Mio. EUR  Anzahl Objekte 10-20 Deutschland bundesweit; jeweiliges Einzugsgebiet > 25.000 Einwohner mit Regionen stabiler Bevölkerungsprognose Risikoprofil Core / Core +",
+        {"entities": [[689, 702, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Fondsvolumen 300 Mio. EUR Zielrendite (IRR) > 6,0 % p. a. Ausschuttung >5,0 % p. a.  Ankaufszeitraum 2024-2026 Laufzeit 31.12.2036 Mindestanlage 10 Mio. EUR Anlageausschuss Ja, entscheidet u. a. über Objekterwerb (Mitglied kann ab 20 Mio. EUR gestellt werden) Gebührenstruktur Marktüblich (auf Anfrage) Projektentwicklungen keine Forward-Deals Möglich, maximal 18 Monate Vorlauf; keine Projektentwicklungsrisiken beim Fonds Erbbaurechte Möglich, sofern Laufzeit > 60 Jahre und angemessene Entschädigung bei Ablauf und Heimfall Status Objektpipeline vorhanden: siehe Folie 16 ff.",
+        {"entities": [[44, 57, "RENDITE"], [71, 83, "AUSSCHÜTTUNGSRENDITE"], [120, 130, "LAUFZEIT"]]},
+    ),
+    (
+        "Niederlande (max. 35 %) Länderallokation Frankreich (max. 35 %) (in % vom Zielvolumen) Skandinavien (Schweden, Dänemark) (max. 35 %) Deutschland (<= 10 %)",
+        {"entities": [[0, 23, "LÄNDERALLOKATION"], [41, 63, "LÄNDERALLOKATION"], [87, 132, "LÄNDERALLOKATION"],  [133, 154, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Führender Immobilien-Investmentmanager in den Nordics für globale ll institutionelle Investoren in Value Add und Core Strategien",
+        {"entities": [[101, 119, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core und Core+ Fonds",
+        {"entities": [[0, 14, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Risikoprofil Core / Core+",
+        {"entities": [[13, 25, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Durchschnittlich geplante jährliche Ausschüttung von 4,5-5,5% auf das investierte Eigenkapital an die Anleger Geplante Gesamtrendite von 5-6% (IRR) auf das eingezahlte Eigenkapital",
+        {"entities": [[54, 62, "AUSSCHÜTTUNGSRENDITE"], [138, 142, "RENDITE"]]},
+    ),
+    (
+        "Geografische Zielallokation nach Investitionsphase des Fonds: 1) Schweden 20-60% Allokation Länder 2) Finnland 20-60% 3) Norwegen 10-40% 4) Dänemark 10-40%",
+        {"entities": [[65, 80, "LÄNDERALLOKATION"], [102, 117, "LÄNDERALLOKATION"], [121, 136, "LÄNDERALLOKATION"], [140, 155, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Deutsches Spezial-Sondervermögen mit festen Anlagebedingungen ($284 KAGB) Immobilien- oder Infrastrukturquote (nach Solvency II) Core / Core+ Euro Hauptstadtregionen und andere Großstädte in den Nordics €500 Mio. 4,5-5,5% 15 Jahre; Fonds hat unbegrenzte Laufzeit; Investmentphase 4 Jahre Maximaler Fremdkapitalanteil 50% (LTV-Ziel bei Ankauf), Langfristiges LTV-Ziel auf Fondsebene ist 45% 0,625% p. a. des Bruttofondsvermögens Zeichnungen ab € 30 Mio. - 0,03 % Rabatt Zeichnungen ab € 50 Mio. - zusatzl. 0,03 % Rabatt 1,1% des Verkehrswertes 0,6% der Bruttoverkaufswert 10% wenn Hurdle Rate 5,0 % p. a. (IRR netto) überschritten wird (nach 15 Jahren berechnet) Ja",
+        {"entities": [[129, 141, "RISIKOPROFIL"], [213, 221, "ZIELRENDITE"], [242, 262, "LAUFZEIT"]]},
+    ),
+    (
+        "Standort Helsinki, Finnland Sektor Bildungswesen, Schule& Kindertagesstätte Vermietbare Fläche 3.321 m? Leerstand bei Ankauf 0% / 0% Ankaufspreis+ Investitionen €21,4 Mio. + €0,2 Mio Eigenkapital €21,6 Mio. Ankaufs- / Stabilisierungs- / Exitrendite 5,0%/ 5,5%/ 5,0% NOI zum Ankaufszeitpunkt / Exit-NOI €1.1m/ €1.2m Zielrenditen (netto für LPs) 5,4% IRR/ 1.5x EM / DY 4,3% Ankauf / Exit Dezember 2023/ Dezember 2033",
+        {"entities": [[345, 349, "ZIELRENDITE"]]},
+    ),
+    (
+        "Evergreen/offene Fondsstrukturenv Core / Core+ Strategien",
+        {"entities": [[34, 46, "RISIKOPROFIL"]]},
+    ),
+    (
+        "BEE Henderson German 2012 Logistik Core/D/Art. 8 € 336 Mio. 12 (voll investiert) 13,0 % p.a.",
+        {"entities": [[35, 39, "RISIKOPROFIL"], [81, 87, "RENDITE"]]},
+    ),
+    (
+        "ICF German Logistics 2014 Logistik Core/D/Art. 8 € 400 Mio. 16 (voll investiert) 12,0 % p.a.",
+        {"entities": [[35, 39, "RISIKOPROFIL"], [81, 87, "RENDITE"]]},
+    ),
+    (
+        "Individualmandat 2015 Logistik Core / D+AU/ ArTt. 6 € 200 Mio. 8 (realisiert) 8,0 % p.a.",
+        {"entities": [[31, 35, "RISIKOPROFIL"], [78, 83, "RENDITE"]]},
+    ),
+    (
+        "European Logistics Partnership” 2017 Logistik Value-Add / Europ/a - € 1.000 Mio. 28 (realisiert) 20,0 % p.a.",
+        {"entities": [[46, 55, "RISIKOPROFIL"], [97, 103, "RENDITE"]]},
+    ),
+    (
+        "European Core Logistics Fund (ECLF 1) 2021 Logistik Core / Euro/p Arat. 8 € 314 Mio. 12 (voll investiert) 7,50 % p.a.",
+        {"entities": [[9, 13, "RISIKOPROFIL"], [106, 112, "RENDITE"]]},
+    ),
+    (
+        "P-Logistik Europa Fonds (ECLF 2) 2022 Logistik Core / Euro/p Arat. 8 € 150 Mio.? A (voll investiert) 6,5 % p.a.?",
+        {"entities": [[47, 51, "RISIKOPROFIL"], [101, 106, "RENDITE"]]},
+    ),
+    (
+        "First Business Parks 2015 Light Industrial Value Add / D+AUT € 100 Mio. 6 (realisiert) 16,0 % p.a.",
+        {"entities": [[43, 52, "RISIKOPROFIL"], [87, 93, "RENDITE"]]},
+    ),
+    (
+        "Unternehmensimmobilien Club 1 2016 Light Industrial Core+/D € 186 Mio. 9 (voll investiert) 13,0 % p.a.",
+        {"entities": [[91, 97, "RENDITE"]]},
+    ),
+    (
+        "Unternehmensimmobilien Club 1 2016 Light Industrial Core+/D € 186 Mio. 9 (voll investiert) 13,0 % p.a.",
+        {"entities": [[52, 57, "RISIKOPROFIL"], [91, 97, "RENDITE"]]},
+    ),
+    (
+        "Unternehmensimmobilien Club 2 2021 Light Industrial Core+/D € 262 Mio. 12 (voll investiert) 9,00 % p.a.",
+        {"entities": [[52, 57, "RISIKOPROFIL"], [92, 98, "RENDITE"]]},
+    ),
+    (
+        "Individualmandat 2022 Light Industrial Value-Add / Nordics € 100 Mio. 5 (voll investiert) 18,0 % p.a.",
+        {"entities": [[39, 48, "RISIKOPROFIL"], [90, 96, "RENDITE"]]},
+    ),
+    (
+        "EUROPEAN CORE LOGISTICS FUND 3",
+        {"entities": [[9, 13, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Core Investitionen",
+        {"entities": [[0, 4, "RISIKOPROFIL"]]},
+    ),
+    (
+        "8 % IRR",
+        {"entities": [[0, 3, "RENDITE"]]},
+    ),
+    (
+        "Rendite-Risiko-Profil Core ° Geographischer Fokus Kontinentaleuropaische Kernvolkswirtschaften nach Allokationsprofil * Sektoraler Fokus Logistikimmobilien nach Allokationsprofil Kapitalstruktur ° Eigenkapital € 250 Mio. ° Fremdkapital 50 % angestrebt, max. 60 % der Immobilienwerte (Objektebene) °e Mindestzeichnung € 10 Mio. Vehikelstruktur ° Rechtsform Immobilien-Spezial-AlF mit festen Anlagebedingungen nach 3 284 KAGB ° Klassifikation Artikel 8 Offenlegungsverordnung ¢ Anlagehorizont 10 Jahre mit Verlängerungsoption um 2 Jahre! ° Geplante Auflage 01 2025 Performanceziel? ° Ausschüttung 6,0 % p.a. (Durchschnitt 10 Jahre Haltedauer) ° Interner Zinsfuß (IRR) 8,0 % p.a. (10 Jahre Haltedauer, Target-IRR)",
+        {"entities": [[22, 26, "RISIKOPROFIL"], [596, 601, "AUSSCHÜTTUNGSRENDITE"], [667, 672, "RENDITE"]]},
+    ),
+    (
+        "Core/Core+, mit Cash-Flow-Stabilität",
+        {"entities": [[0, 10, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Zielausschüttung: min. 5,10%",
+        {"entities": [[24, 29, "ZIELAUSSCHÜTTUNG"]]},
+    ),
+    (
+        "Zielrendite (IRR): min. 5,50%",
+        {"entities": [[24, 29, "ZIELRENDITE"]]},
+    ),
+    (
+        "Rewe & Lidl Maxhütte-Haidhof é ae: 6 s Bahnhof Ankermieter REWE & Lidl er WALT 20 und 17 Jahre Miete p.a. 1.127.916 € Kaufpreis 21,43 Mio. € Faktor 19,00 x LTV / Zins 80% / 4,0% Ausschüttung 5,7 % IRR 7,1%",
+        {"entities": [[193, 198, "AUSSCHÜTTUNGSRENDITE"], [203, 207, "ZIELRENDITE"]]},
+    ),
+    (
+        "Real Estate Prime Europe Access the Core of European Prime Cities with a green SRI fund including a genuine low carbon commitment",
+        {"entities": [[36, 40, "RISIKOPROFIL"]]},
+    ),
+    (
+        "(FR, UK, DE, BE, NL, LU, Nordics, Allocation SP, IT, CH)",
+        {"entities": [[1, 32, "LÄNDERALLOKATION"], [45, 55, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "IRR: 6% - 7%",
+        {"entities": [[5, 12, "RENDITE"]]},
+    ),
+    (
+        "Europe | Germany 67 Value Add",
+        {"entities": [[9, 16, "LÄNDERALLOKATION"], [20, 29, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Germany, Norway 336 Core Plus",
+        {"entities": [[0, 7, "LÄNDERALLOKATION"], [20, 29, "RISIKOPROFIL"]]},
+    ),
+    (
+        "UK",
+        {"entities": [[0, 2, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "NORWAY",
+        {"entities": [[0, 6, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "9.8% IRR",
+        {"entities": [[0, 4, "RENDITE"]]},
+    ),
+    (
+        "Investment volume down 52% to €2.3 billion, with 4,000 100 14% value-add and core-plus increasing YoY",
+        {"entities": [[63, 86, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Geared Gross IRR‘ seeking a range of 16-18% per annum",
+        {"entities": [[37, 43, "RENDITE"]]},
+    ),
+    (
+        "Open-ended fund 24 months, incl. rolling reinvestment Sale of individual assets with respective management contracts or geared leases IRR: >6.5% | CoC: >5.0%",
+        {"entities": [[0, 10, "LAUFZEIT"], [139, 144, "RENDITE"]]},
+    ),
+    (
+        "Our investment strategy focuses on investing in upscale hotels in European prime locations, including DACH, Italy, Spain, Portugal, France, UK, Denmark, Benelux,and Poland.",
+        {"entities": [[102, 171, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Core+ assets with value-add potential, Emerging Gateway Cities Helsinki] Core+ with Value well-mitigated risk and great upside Potential potential through asset improvement or = Max. 20% UK & Ireland {no contract renegotiation currency risk hedging], 80% tinental E > IRR target of 6-9%",
+        {"entities": [[0, 5, "RISIKOPROFIL"], [282, 286, "RENDITE"]]},
+    ),
+    (
+        "10% net IRR since inception in 2018?",
+        {"entities": [[0, 3, "RENDITE"]]},
+    ),
+    (
+        "Eurozone: Benelux, France and Germany",
+        {"entities": [[10, 37, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Open-ended, with quarterly liquidity (redemption rights, dual pricing)",
+        {"entities": [[0, 10, "LAUFZEIT"]]},
+    ),
+    (
+        "Class A & B (Institutional): 0.93% on NAV; Class D (Wholesale): 1.80% on NAV; Class P (Wholesale): 1.25% on NAV",
+        {"entities": [[29, 34, "MANAGMENTGEBÜHREN"], [64, 69, "MANAGMENTGEBÜHREN"], [99, 104, "MANAGMENTGEBÜHREN"]]},
+    ),
+    (
+        "Risk profile: favour core > © at least and core+ assets with a targeted N 2 n allocation to value add assets to enhance returns",
+        {"entities": [[21, 25, "RISIKOPROFIL"], [43, 48, "RISIKOPROFIL"]]},
+    ),
+    (
+        "The Netherlands (38 assets) = Germany (9 assets) 10 largest Country assets split France (8 assets)",
+        {"entities": [[0, 15, "LÄNDERALLOKATION"], [30, 37, "LÄNDERALLOKATION"], [81, 87, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Expected IRR 10.9%",
+        {"entities": [[13, 18, "ZIELRENDITE"]]},
+    ),
+    (
+        "Structure Open-end, perpetual life, Luxembourg domiciled Initial Target Size* €2 billion 6-8% total return,",
+        {"entities": [[10, 18, "LAUFZEIT"], [89, 93, "RENDITE"]]},
+    ),
+    (
+        "Geographic Focus: UK, Ireland, Iberia, Nordics, Netherlands, Germany, France, Italy",
+        {"entities": [[18, 83, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "IRR of 13-14%",
+        {"entities": [[7, 13, "RENDITE"]]},
+    ),
+    (
+        "Value-add",
+        {"entities": [[0, 9, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Geographic allocation NORDICS UNITED KINGDOM GERMANY FRANCE PORTUGAL BENELUX",
+        {"entities": [[22, 76, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Strong track record delivering a 17% net IRR, 1.7x net multiple across all divested assets (both discretionary and non-discretionary mandates)",
+        {"entities": [[33, 36, "RENDITE"]]},
+    ),
+    (
+        "Targeting a 7-8% net annual return and a 3-4% dividend yield, reflecting a target LTV of 35% (capped at 37.5%)",
+        {"entities": [[12, 16, "RENDITE"]]},
+    ),
+    (
+        "Sweden Norway Denmark Finland",
+        {"entities": [[0, 29, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Logistics Residential Office Other",
+        {"entities": [[0, 34, "SEKTORENALLOKATION"]]},
+    ),
+    (
+        "Fund Term Open-ended with an initial 24-month lock-in for new investors",
+        {"entities": [[10, 20, "LAUFZEIT"]]},
+    ),
+    (
+        "Management fee of 85 bps on NAV.",
+        {"entities": [[18, 24, "MANAGMENTGEBÜHREN"]]},
+    ),
+    (
+        "Core/Core+ strategy, with tactical exposure to development projects aiming at enhancing the quality of the portfolio over time",
+        {"entities": [[0, 10, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Fund term: Open-ended",
+        {"entities": [[11, 21, "LAUFZEIT"]]},
+    ),
+    (
+        "Return targets: The fund targets a net internal rate of return (IRR) of 8% and a net annual income yield of 5% with planned quarterly distributions’.",
+        {"entities": [[72, 74, "RENDITE"]]},
+    ),
+    (
+        "Geographic scope: The fund has a broad mandate to invest in commercial and residential real estate across Sweden, Denmark, Finland, and Norway. 50% LTV Asset selection: Heirs to acquire high-quality, income-generating properties in major Nordic cities and enhance their value through active asset management. Portfolio construction: The goal is to build diversified portfolios that are appealing to core buyers upon exit.",
+        {"entities": [[106, 142, "LÄNDERALLOKATION"]]},
+    ),
+    (
+        "Experience: Since 2012, | | has demonstrated its capability to build diversified and resilient portfolios for its core-plus funds. German Real Estate Quota advantage . Local expertise: extensive local relationships and proprietary deal flow in key Nordic markets provide a strategic advantage.",
+        {"entities": [[114, 123, "RISIKOPROFIL"]]},
+    ),
+    (
+        "Target returns: 8% net IRR with 5% net annual income yield! * Geographic focus: Sweden, Denmark, Norway and Finland « Target leverage: 50% LTV (excluding short-term borrowing) « Sector exposure: office, logistics, public properties, retail (focused on grocery anchored and necessity driven retail) and residentials « Investment focus: high quality properties,",
+        {"entities": [[16, 18, "RENDITE"], [80, 115, "LÄNDERALLOKATION"], [195, 239, "SEKTORENALLOKATION"]]},
+    ),
+    (
+        "The Fund 2 xemoours common limited partnership (SCS) (SICAV-RAIF) Investment Objective To pursue investments in commercial and residential properties throughout the Nordic Region Fund Target Size €300 million (equity) Return Targets Target net IRR of 8%, target net annual income yield of 5%",
+        {"entities": [[251, 253, "RENDITE"]]},
+    )
+
+]
+
+
+
+

project/backend/spacy-service/spacy_training/training_model.py

@@ -0,0 +1,40 @@
+import os
+from pathlib import Path
+
+import spacy
+from spacy.cli.train import train
+
+from spacy.tokens import DocBin
+
+from tqdm import tqdm
+
+from training_data import TRAINING_DATA
+
+nlp = spacy.blank("de")
+
+# create a DocBin object
+db = DocBin()
+
+for text, annot in tqdm(TRAINING_DATA):
+    doc = nlp.make_doc(text)
+    ents = []
+    # add character indexes
+    for start, end, label in annot["entities"]:
+        span = doc.char_span(start, end, label=label, alignment_mode="contract")
+        if span is None:
+            print(f"Skipping entity: |{text[start:end]}| Start: {start}, End: {end}, Label: {label}")
+        else:
+            ents.append(span)
+    # label the text with the ents
+    doc.ents = ents
+    db.add(doc)
+
+# save the DocBin object
+os.makedirs("./data", exist_ok=True)
+db.to_disk("./data/train.spacy")
+
+config_path = Path("config.cfg")
+output_path = Path("output")
+
+print("Starte Training...")
+train(config_path, output_path)

project/docker-compose.yml

@@ -37,6 +37,8 @@ services:
       retries: 10
     ports:
       - 5050:5000
+    volumes:
+      - ./backend/spacy-service/spacy_training:/app/spacy_training
 
   ocr:
     build:

project/frontend/package.json

@@ -23,6 +23,7 @@
 		"@tanstack/react-router": "^1.114.3",
 		"@tanstack/react-router-devtools": "^1.114.3",
 		"@tanstack/router-plugin": "^1.114.3",
+		"file-saver": "^2.0.5",
 		"react": "^19.0.0",
 		"react-dom": "^19.0.0",
 		"react-material-file-upload": "^0.0.4",

project/frontend/src/components/KPIForm.tsx

@@ -400,14 +400,13 @@ export function KPIForm({ mode, initialData, onSave, onCancel, loading = false,
 function generateSpacyEntries(formData: Partial<Kennzahl>) {
     const label = formData.name?.trim().toUpperCase() || "";
     return (formData.examples || []).map(({ sentence, value }) => {
-        const trimmedValue = value.trim();
+        const start = sentence.indexOf(value);
-        const start = sentence.indexOf(trimmedValue);
         if (start === -1) {
-            throw new Error(`"${trimmedValue}" nicht gefunden in Satz: "${sentence}"`);
+            throw new Error(`"${value}" nicht gefunden in Satz: "${sentence}"`);
         }
         return {
             text: sentence,
-            entities: [[start, start + trimmedValue.length, label]]
+            entities: [[start, start + value.length, label]]
         };
     });
 }

project/frontend/src/routes/config.tsx

@@ -78,29 +78,22 @@ function ConfigPage() {
         }
     };
 
-    const handleTriggerTraining = async () => {
+    const handleTriggerTraining = () => {
         setTrainingRunning(true);
+        setSnackbarMessage("Training wurde gestartet.");
+        setSnackbarOpen(true);
 
-        try {
+        fetch(`${API_HOST}/api/spacy/train`, {
-            const response = await fetch(`${API_HOST}/api/spacy/train`, {
+            method: "POST",
-                method: "POST",
+        }).catch(err => {
-            });
-
-            if (!response.ok) throw new Error("Training konnte nicht gestartet werden");
-
-            // Erfolgsmeldung erst hier anzeigen
-            setSnackbarMessage("Training wurde gestartet.");
-            setSnackbarOpen(true);
-
-            pollTrainingStatus(); // jetzt starten
-        } catch (err) {
-            console.error(err);
             setSnackbarMessage("Fehler beim Starten des Trainings.");
             setSnackbarOpen(true);
-            setTrainingRunning(false);
+            console.error(err);
-        }
+        });
-    };
 
+        pollTrainingStatus(); // Starte Überwachung
+
+    };
 
     const pollTrainingStatus = () => {
         const interval = setInterval(async () => {