diff --git a/hal-classifier/README.md b/hal-classifier/README.md
new file mode 100644
index 0000000..f65d468
--- /dev/null
+++ b/hal-classifier/README.md
@@ -0,0 +1,47 @@
+# hal-classifier
+
+Cette instance propose des outils de classification pour HAL
+
+
+
+## Utilisation
+
+- [v1/classhalen](#v1%2fdetect-lang)
+
+### v1/en/classhalen
+
+Ce service permet d'attribuer un domaine de premier niveau de la base HAL à des documents en anglais
+
+Prend en entrée du JSON avec deux champs: `id` et `value`, et renvoie un JSON
+avec le code langue dans le champ `value`.
+
+### v1/fr/classhalfr
+
+Ce service permet d'attribuer un domaine de premier niveau de la base HAL à des documents en français
+
+Prend en entrée du JSON avec deux champs: `id` et `value`, et renvoie un JSON
+avec le code langue dans le champ `value`.
+
+#### Paramètres de v1/en/classhalen et v1/fr/classhalfr
+
+| nom    | description                                                            |
+| :----- | :--------------------------------------------------------------------- |
+| indent | `true` ou `false`, indente le JSON résultat ou non (`true` par défaut) |
+
+#### Exemple
+
+```bash
+$ cat<<EOF|curl --proxy "" -X POST --data-binary @- "http://hal-classifier.tdmservices.intra.inist.fr/v1/en/classhalen?"
+[{"id":"1","value":"In the southern French Massif Central, the Montagne Noire axial zone is a NE-SW elongated granite-migmatite dome emplaced within Visean south-verging recumbent folds and intruded by syn- to late-migmatization granitoids. The tectonic setting of this dome is still disputed, thus several models have been proposed. In order to better understand the emplacement mechanism of this dome, petrofabric and Anisotropy of Magnetic Susceptibility (AMS) studies have been carried out. In the granites and migmatites that form the dome core, magmatic texture and to a lesser extent weak solid-state texture are dominant. As a paramagnetic mineral, biotite is the main carrier of the magnetic susceptibility. On the basis of 135 AMS sites, the magnetic fabrics appear as independent of the lithology but related to the dome architecture. Coupling our results with previous structural and geochronological studies, allows us to propose a new emplacement model. Between 340-325 Ma, the Palaeozoic series underwent a compressional deformation represented by nappes and recumbent folds involving the thermal event leading to partial melting. Until ~325-310 Ma, the dome emplacement was assisted by diapiric processes. An extensional event took place at 300 Ma, after the emplacement of the late to post-migmatitic granitic plutons. In the northeast side of the dome, a brittle normal-dextral faulting controlled the opening of the Graissessac coal-basin."}]
+EOF
+
+```
+
+Sortie
+
+```json
+[{
+    "id": 1,
+    "value": "Planète et Univers [physics]"
+}]
+```
diff --git a/hal-classifier/V1/en/HALen_matrixKDT.npy b/hal-classifier/V1/en/HALen_matrixKDT.npy
new file mode 100644
index 0000000..b1759e2
--- /dev/null
+++ b/hal-classifier/V1/en/HALen_matrixKDT.npy
Binary files differ
diff --git a/hal-classifier/V1/en/classhalen.ini b/hal-classifier/V1/en/classhalen.ini
new file mode 100644
index 0000000..7546967
--- /dev/null
+++ b/hal-classifier/V1/en/classhalen.ini
@@ -0,0 +1,41 @@
+# OpenAPI Documentation - JSON format (dot notation)
+mimeType = application/json
+post.operationId = post-v1-en-classhalen
+post.description = Calcule la classe de premier niveau de la base HAL pour des documents en anglais et renvoie la verbalisation de la classe
+post.responses.default.description = Renvoie un Json composé de `id`, `value` avec `value` la verbalisation en français de la classe
+post.responses.default.content.application/json.schema.$ref =  #/components/schemas/JSONStream
+post.summary = Classification HAL
+post.tags.0 = Classification supervisée
+post.requestBody.required = true
+post.requestBody.content.application/json.schema.$ref = #/components/schemas/JSONStream
+post.parameters.0.in = query
+post.parameters.0.name = path
+post.parameters.0.schema.type = string
+post.parameters.0.description = The path in each object to enrich with an Python script
+post.parameters.1.in = query
+post.parameters.1.name = indent
+post.parameters.1.schema.type = boolean
+post.parameters.1.description = Indent or not the JSON Result
+
+[use]
+plugin = @ezs/spawn
+plugin = @ezs/basics
+plugin = @ezs/storage
+plugin = @ezs/analytics
+
+[JSONParse]
+#legacy = false
+separator = *
+
+[expand]
+path = env('path', 'value')
+size = 100
+# in production mode, uncomment the following line
+# cache = boost
+
+[expand/exec]
+# command should be executable !
+command = ./dist_kn3_HAL_en_v1.py
+
+[dump]
+indent = env('indent', false)
diff --git a/hal-classifier/V1/en/dico_nd.pkl b/hal-classifier/V1/en/dico_nd.pkl
new file mode 100644
index 0000000..a4c187a
--- /dev/null
+++ b/hal-classifier/V1/en/dico_nd.pkl
Binary files differ
diff --git a/hal-classifier/V1/en/dist_kn3_HAL_en_v1.py b/hal-classifier/V1/en/dist_kn3_HAL_en_v1.py
new file mode 100644
index 0000000..52aad0f
--- /dev/null
+++ b/hal-classifier/V1/en/dist_kn3_HAL_en_v1.py
@@ -0,0 +1,98 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Created on Thu Aug 12 10:19:31 2021
+
+@author: cuxac
+"""
+
+
+
+
+from scipy.spatial import cKDTree as KDTree
+from scipy.special import softmax
+import numpy as np
+import sys
+import json
+from collections import defaultdict#,Counter
+import operator
+from more_itertools import locate
+import statistics
+import pickle
+import fasttext
+
+
+
+
+model=fasttext.load_model("modelhal0EN2.bin")
+D=np.load("HALen_matrixKDT.npy")
+
+dico_nd_pkl=open("dico_nd.pkl","rb")    
+dico_nd=pickle.load(dico_nd_pkl)
+
+
+
+verb_class={"spi":"Sciences de l'ingénieur [physics]","shs":"Sciences de l'Homme et Société","sdv":"Sciences du Vivant [q-bio]","sdu":"Planète et Univers [physics]","sde":"Sciences de l'environnement","scco":"Sciences cognitives","phys":"Physique [physics]","nlin":"Science non linéaire [physics]","math":"Mathématiques [math]","info":"Informatique [cs]","chim":"Chimie","stat":"Statistiques","qfin":"Économie et finance quantitative [q-fin]"}
+
+kdtree=KDTree(D)
+
+
+distlist=[]
+list_defis=[]
+
+n=0
+K=50
+
+for line in sys.stdin:
+    data = json.loads(line)
+    text=data['value']
+
+    mv=model.get_sentence_vector(text.strip())
+
+    ppv=kdtree.query(mv,k=K,p=2)#,distance_upper_bound=0.05)
+    dmax_k=kdtree.query(mv,k=[K],p=2)[0][0]
+    dN_ppv=ppv[0]/dmax_k
+    distlist.append(ppv[0][0]/dmax_k)
+ 
+    list_defis=list(ppv[1])
+    list_defis_label=list(dico_nd[i] for i in list_defis)
+    
+    r=zip(list_defis_label,list(ppv[0]))
+    dis=defaultdict(list)
+    for i in set(r):
+        dis[i[0]].append(1/i[1])
+    for k in dis.keys():     
+        dis[k]=sum(dis[k])
+    ddd=np.array(list(dis.values()))
+    sm=softmax(ddd)
+    lab=list(dis.keys())
+    res=zip(lab,sm)
+    d4=dict(res)
+    classmax=max(d4.items(),key=operator.itemgetter(1))[0]
+    cmax=classmax.split('_')[1].split('.')[0]
+    
+
+    if ppv[0][0]<100:
+
+        mm=max(list_defis_label,key=list_defis_label.count)
+
+        ind=list(locate(list_defis_label, lambda a: a ==mm))
+        indd=list(dN_ppv)
+        dmax=[indd[i] for i in ind]
+        dist_mean=statistics.mean(dmax)
+        prob0=round((len(ind))/(K+len(set(list_defis_label))),3)
+        prob=d4[classmax]
+        if len(ind)<25:
+            dist_mean=str(round(dist_mean,3))+' / '+str(len(ind))
+        
+        data['value']=verb_class[cmax]#,prob
+        sys.stdout.write(json.dumps(data))
+        sys.stdout.write('\n')
+        
+        n+=1
+
+
+
+
+
+
diff --git a/hal-classifier/V1/en/modelhal0EN2.bin b/hal-classifier/V1/en/modelhal0EN2.bin
new file mode 100644
index 0000000..1a27209
--- /dev/null
+++ b/hal-classifier/V1/en/modelhal0EN2.bin
Binary files differ
diff --git a/hal-classifier/examples.http b/hal-classifier/examples.http
new file mode 100644
index 0000000..25910e5
--- /dev/null
+++ b/hal-classifier/examples.http
@@ -0,0 +1,15 @@
+# These examples can be used directly in VSCode, using REST Client extension (humao.rest-client)
+
+# Classification dans les domaines de niveau de la base HAL
+POST https://hal-classifier.services.inist.fr/v1/en/classhalen?indent=true HTTP/1.1
+Content-Type: application/json
+
+[
+{
+"id":1,
+"value":"In the southern French Massif Central, the Montagne Noire axial zone is a NE-SW elongated granite-migmatite dome emplaced within Visean south-verging recumbent folds and intruded by syn- to late-migmatization granitoids. The tectonic setting of this dome is still disputed, thus several models have been proposed. In order to better understand the emplacement mechanism of this dome, petrofabric and Anisotropy of Magnetic Susceptibility (AMS) studies have been carried out. In the granites and migmatites that form the dome core, magmatic texture and to a lesser extent weak solid-state texture are dominant. As a paramagnetic mineral, biotite is the main carrier of the magnetic susceptibility. On the basis of 135 AMS sites, the magnetic fabrics appear as independent of the lithology but related to the dome architecture. Coupling our results with previous structural and geochronological studies, allows us to propose a new emplacement model. Between 340-325 Ma, the Palaeozoic series underwent a compressional deformation represented by nappes and recumbent folds involving the thermal event leading to partial melting. Until ~325-310 Ma, the dome emplacement was assisted by diapiric processes. An extensional event took place at 300 Ma, after the emplacement of the late to post-migmatitic granitic plutons. In the northeast side of the dome, a brittle normal-dextral faulting controlled the opening of the Graissessac coal-basin."
+},
+{"id":2,
+"value":"The COVID-19 pandemic, also known as the coronavirus pandemic, is an ongoing global pandemic of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus2 (SARS-CoV-2). It was first identified in December 2019 in Wuhan, China. The World Health Organization declared the outbreak a Public Health Emergency of International Concern on 20 January 2020, and later a pandemic on 11 March 2020. As of 2 April 2021, more than 129 million cases have been confirmed, with more than 2.82 million deaths attributed to COVID-19, making it one of the deadliest pandemics in history."
+}
+]
diff --git a/hal-classifier/requirements.txt b/hal-classifier/requirements.txt
new file mode 100644
index 0000000..c2d846a
--- /dev/null
+++ b/hal-classifier/requirements.txt
@@ -0,0 +1,4 @@
+more-itertools
+numpy
+scipy
+fasttext
diff --git a/hal-classifier/swagger.json b/hal-classifier/swagger.json
new file mode 100644
index 0000000..436f673
--- /dev/null
+++ b/hal-classifier/swagger.json
@@ -0,0 +1,20 @@
+{
+    "servers": [{
+        "url": "{scheme}://{hostname}",
+        "variables": {
+            "scheme": {
+                "default": "https"
+            }
+        }
+    }],
+    "tags": [
+        {
+            "name": "Classification supervisée",
+            "description": "Classification supervisée dans les domaines de HAL",
+            "externalDocs": {
+                "description": "Plus de documentation",
+                "url": "https://gitbucket.inist.fr/tdm/web-services/tree/master/hal-classifier"
+            }
+        }
+    ]
+}