Same architecture and training data as 81melody/algerianDeBERTa-realestate-ner (the precision-optimised),This one trades a few precision points for measurably higher recall, especially on neighbourhood names

Handles the exact language mix found on Facebook Marketplace and Algerian classified groups: Darja, Arabizi, French, and heavy code-switching.

v1 vs v2 — when to use which

	This model v2	v1
Val F1	0.9780	0.9858
Val Precision	0.9646	0.9784
Val Recall	0.9918	0.9933
Avg entities / listing (prod)	8.45	7.84
NEIGHBORHOOD / 100 texts	72	48
Avg entity confidence	0.903	0.931

Use v2 when:

You need maximum location coverage , it finds ~50% more neighbourhood mentions per listing ( "Hussein dey", "Bab Ezzouar", "l3achour" are consistently extracted even when phrasing is irregular)
You are building a search / matching engine and prefer recall over precision (missing a match is worse than a false positive)
You are running a second-pass re-extraction to find locations missed by a stricter model

Use v1 when:

You need clean, high-confidence entities for structured storage or analytics
Precision matters more than recall (e.g. price indexing, surface-area statistics)

Quick comparaison

Summary comparaison :

Score distribution

Entity counts

Model Highlights


Architecture	DeBERTa-v2 — 12 layers, hidden=512, 8 heads, 2048 FFN
Base model	algerianDeBERTa (pre-trained on Algerian web text)
Task	Token classification — 27 BIO labels, 13 entity types
Languages	Algerian Darja · Arabizi · French · MSA · Code-switched
Domain	Real estate classifieds (sales, rentals, land, villas, apartments)
Val F1	0.9780
Val Recall	0.9918
Parameters	~60M
License	Apache 2.0

No separate test-set evaluation was run for this version. For test metrics (F1 = 0.9672 on a held-out set of 95 posts) see v1 model card.

Quick Start

Option 1 : `pipeline` (standard, recommended for short texts)

Uses aggregation_strategy="max": for each surface word the subword token with the highest entity-class score wins, then consecutive spans that have the same type are merged automatically

python

from transformers import pipeline

ner = pipeline(
    "token-classification",
    model="81melody/algerianDeBERTa-realestate-ner-v2",
    aggregation_strategy="max",
)


print(ner("سلام ، خصني اف2 فالعاصمة في ميسوني ولا اودان ولا ديدوش ، في هاد الجويه لي عندو يتوصل معيا في الخاص"))

print(ner("Appartement F4 à vendre Oran centre 120m² 4ème étage acte notarié"))

For texts that may exceed 192 tokens, pass sliding-window arguments directly to the pipeline call:

python

result = ner(
    long_text,
    truncation=True,
    max_length=192,
    stride=64,
)

Each result dict contains entity_group, word, score, start, end.

Option 2 : Manual sliding window inference (production / long texts)

For real estate posts that frequently exceed one chunk + to adapt with The small vocab size of the first version of the base model (30k), the approach below is more robust for production use: it averages the probability vectors of overlapping tokens across all chunks, then merges subword pieces that form the same surface word before BIO decoding

This fixes a common artefact where words like "cherche" (tokenised as ["cher", "che"]) or prices like "1.700" (tokenised as ["1", ".", "700"]) get truncated mid-word if a trailing subword happens to predict O

python

from transformers import AutoTokenizer, AutoModelForTokenClassification
import torch
import numpy as np
from typing import List

MODEL_NAME  = "81melody/algerianDeBERTa-realestate-ner-v2"
MAX_SEQ_LEN = 192
STRIDE      = 64

tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model     = AutoModelForTokenClassification.from_pretrained(MODEL_NAME)
model.eval()
device    = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)


def extract_entities(text: str) -> List[dict]:
   
    enc = tokenizer(
        text,
        return_tensors="pt",
        max_length=MAX_SEQ_LEN,
        stride=STRIDE,
        truncation=True,
        return_overflowing_tokens=True,
        return_offsets_mapping=True,
        padding="max_length",
    )
    enc.pop("overflow_to_sample_mapping", None)
    offsets = enc.pop("offset_mapping")          


    with torch.no_grad():
        logits = model(**{k: v.to(device) for k, v in enc.items()}).logits
    probs     = torch.softmax(logits, dim=-1).cpu().numpy()  
    attention = enc["attention_mask"].numpy()                
    off_np    = offsets.numpy()                             


    char_probs = {}
    for c in range(probs.shape[0]):
        for t in range(off_np.shape[1]):
            if attention[c, t] == 0:
                continue
            cs, ce = int(off_np[c, t, 0]), int(off_np[c, t, 1])
            if cs == 0 and ce == 0:         
                continue
            if cs not in char_probs:
                char_probs[cs] = {"end": ce, "vecs": [probs[c, t]]}
            else:
                char_probs[cs]["vecs"].append(probs[c, t])

    if not char_probs:
        return []

    items = sorted(char_probs.items())

    word_tokens = []
    w_start, w_info = items[0]
    w_end   = w_info["end"]
    w_avg_p = np.mean(w_info["vecs"], axis=0)  

    for cs, info in items[1:]:
        if cs == w_end:                         
            w_end = info["end"]
        else:                                    
            word_tokens.append({"start": w_start, "end": w_end, "avg_p": w_avg_p})
            w_start = cs
            w_end   = info["end"]
            w_avg_p = np.mean(info["vecs"], axis=0)
    word_tokens.append({"start": w_start, "end": w_end, "avg_p": w_avg_p})

    label_map = model.config.id2label
    entities, current = [], None

    for w in word_tokens:
        idx   = int(np.argmax(w["avg_p"]))
        label = label_map[idx]
        score = float(w["avg_p"][idx])

        if label == "O":
            if current:
                entities.append(current)
                current = None

        elif label.startswith("B-"):
            if current:
                entities.append(current)
            current = {
                "entity": label[2:],
                "word":   text[w["start"]:w["end"]],
                "score":  score,
                "_sc": [score], "_s": w["start"], "_e": w["end"],
            }

        elif label.startswith("I-"):
            etype = label[2:]
            if current and current["entity"] == etype:
                current["word"]  = text[current["_s"]:w["end"]]
                current["_e"]    = w["end"]
                current["_sc"].append(score)
                current["score"] = float(np.mean(current["_sc"]))
            else:
                if current:
                    entities.append(current)
                current = {
                    "entity": etype,
                    "word":   text[w["start"]:w["end"]],
                    "score":  score,
                    "_sc": [score], "_s": w["start"], "_e": w["end"],
                }

    if current:
        entities.append(current)

    return [
        {"entity": e["entity"], "word": e["word"], "score": round(e["score"], 6)}
        for e in entities
    ]
print(extract_entities('khasni appartement fi bab zouar 200 m2'))
#[{'entity': 'TRANSACTION', 'word': 'khasni', 'score': 0.703333},
#{'entity': 'PROPERTY_TYPE', 'word': 'appartement', 'score': 0.973926},
#{'entity': 'NEIGHBORHOOD', 'word': 'bab zouar', 'score': 0.902971},
#{'entity': 'SURFACE', 'word': '200 m2', 'score': 0.930045}]

Entity Schema

The model uses a 27-label BIO scheme covering 13 entity types.

Entity	Description	Algerian Examples
`PROPERTY_TYPE`	Category of asset	`شقة` · `villa` · `appartement` · `terrain` · `carcasse` · `haouch`
`APT_CLASS`	Apartment layout	`F2` · `F3` · `F4` · `F5` · `S+1` · `Studio`
`TRANSACTION`	Listing intent	`للبيع` · `location` · `louer` · `خاصني` · `echange`
`WILAYA`	Algerian province (all 48)	`Alger` · `Oran` · `Constantine` · `16` · `ولاية وهران`
`CITY`	City / commune	`Bab Ezzouar` · `Sidi Yahia` · `Ain Benian`
`NEIGHBORHOOD`	Quarter / district / street	`باب الزوار` · `Hydra` · `Hai Yasmine` · `Télemly`
`PRICE`	Price in DZD, DA, or slang	`8 500 000 da` · `12M` · `950 000` · `1.5 milliards` · `800 U`
`SURFACE`	Area in m², metres, hectares	`90m²` · `120 mètres` · `85 متر` · `2 hectares`
`FLOOR`	Floor level	`3ème étage` · `الطابق الثالث` · `RDC` · `R+2`
`PHONE`	Contact number (anonymized)	`[PHONE]`
`AMENITY`	Features / utilities	`garage` · `مصعد` · `piscine` · `jardin` · `بيدون`
`DOCUMENT`	Legal papers	`عقد` · `livret foncier` · `AADL` · `acte notarié` · `timbre`
`CONDITION`	Property state	`neuf` · `rénové` · `قديم` · `semi-fini` · `en construction`

Performance

Evaluated on the validation set (87 posts / 586 tokenized chunks).

Val F1        (micro, seqeval):  0.9780
Val Precision (micro):           0.9646
Val Recall    (micro):           0.9918

Test-set evaluation (strict entity-level seqeval on 95 held-out posts) was performed only on the last epoch (test F1 = 0.9672), This version was not re-evaluated on the test set to avoid data leakage into version selection

Production extraction profile (2,505 unseen listings)

Measured on posts never seen during training:

Metric	v2	v1
Avg entities / listing	8.45	7.84
Avg entity confidence	0.903	0.931
Low-conf entities (<0.80)	3,154	1,537
NEIGHBORHOOD / 100 texts	72.3	48.2
SURFACE / 100 texts	73.1	67.3
PRICE / 100 texts	48.1	44.3

The +24 NEIGHBORHOOD gap is the strongest production signal for choosing this version: it reliably spans multi-word neighbourhood names ("Hussein dey", "Bab Ezzouar", "bordj el kiffan") that the more conservative v1 sometimes under-segments

Training Details

yaml

base_model:        algerianDeBERTa (DeBERTa-v2)
architecture:      DebertaV2ForTokenClassification
num_labels:        27  (BIO, 13 entity types)
saved_at_epoch:    7   


max_seq_len:       192
stride:            64


optimizer:         AdamW
peak_lr:           2e-5
llrd_factor:       0.9
weight_decay:      0.01
grad_accum_steps:  2     

warmup_ratio:      0.1
schedule:          cosine with warmup


label_smoothing:   0.05
class_weighting:   inverse-frequency, capped at 10×
dropout:           0.1

Limitations

Lower precision than v1: Confidence scores are ~2–5 pp lower across all entity types. Surface-only units (m, متر without a number) and single-character fragments (ر, ف) can appear as false-positive WILAYA or FLOOR entities. A confidence threshold of score ≥ 0.75 removes the bulk of these
NEIGHBORHOOD F1: Neighbourhood names in Algeria vary widely in spelling. This version finds more of them but also introduces more borderline extractions
PRICE splitting: Multi-part prices (6 000 000) are occasionally split into separate entities rather than merged as one span.
Geography scope: Optimised for the 58 Algerian wilayas
Platform distribution: Trained on Facebook posts , casual, informal register

Intended Use

Use case	Notes
Location-aware search / matching	Best use case for this version maximises neighbourhood recall
Lead enrichment pipelines	Use when missing a location is worse than a noisy one
Candidate generation + reranking	Run epoch 7 to cast wide net, rerank with v1 scores
Training data generation	Higher recall produces more silver labels for rare location types

Citation

If you use this model or the dataset in your research, please cite:

bibtex

@misc{himeur2026algeriandeberta_ner_v2,
  title        = {algerianDeBERTa-realestate-ner-v2: Named Entity Recognition
                  for Algerian Real Estate Text in Darja, Arabizi, and French},
  author       = {Himeur, Ayoub},
  year         = {2026},
  publisher    = {Hugging Face},
  url          = {https://huggingface.co/81melody/algerianDeBERTa-realestate-ner-v2},
  note         = {Fine-tuned DeBERTa-v2 on annotated
                  Algerian Facebook real estate posts, 13 entity types}
}

License

Apache 2.0 free for commercial and research use with attribution

Contact

mohamed.himeur@student.unamur.be

81melody/algerianDeBERTa-realestate-ner-v2

algerianDeBERTa-realestate-ner-v2

v1 vs v2 — when to use which

Quick comparaison

Model Highlights

Quick Start

Option 1 : pipeline (standard, recommended for short texts)

Option 2 : Manual sliding window inference (production / long texts)

Entity Schema

Performance

Production extraction profile (2,505 unseen listings)

Training Details

Limitations

Intended Use

Citation

License

Contact

81melody/
algerianDeBERTa-realestate-ner-v2

Option 1 : `pipeline` (standard, recommended for short texts)