๐ฑ๐ฐ Sri Lanka Location Intelligence System
Find every location, business, and place in Sri Lanka using AI-powered semantic search.
A comprehensive geospatial intelligence system covering 117,446 locations across Sri Lanka โ including police stations, hospitals, schools, places of worship, businesses, roads, buildings, natural features, and more. Powered by multilingual-e5-large (1024-dim embeddings, 100 languages) with FAISS IVFFlat vector search for instant semantic lookup.
Model Details
| Property |
Value |
| Base Model |
intfloat/multilingual-e5-large |
| Architecture |
XLMRobertaModel (24 layers, 16 heads) |
| Embedding Dimension |
1024 |
| Model Size |
2.2 GB (SafeTensors) |
| Max Sequence Length |
512 tokens |
| Languages |
100 languages (English, Sinhala, Tamil, Hindi, Arabic, Chinese, Japanese, Korean, French, German, Spanish, and 90+ more) |
| License |
MIT (model) + ODbL (location data) |
Repository Contents
| File |
Size |
Description |
model.safetensors |
2.2 GB |
multilingual-e5-large model weights |
models/ |
โ |
Tokenizer, config, and sentence-transformers config |
srilanka_places.db |
38 MB |
SQLite database of 117,446 Sri Lanka locations |
srilanka_embeddings_large.npy |
~480 MB |
Pre-computed 1024-dim embeddings (117,446 x 1024) |
faiss_index_large.bin |
~480 MB |
FAISS IVFFlat index (nlist=200) for instant semantic search |
id_mapping_large.npy |
~2 MB |
ID-to-index mapping for result retrieval |
metadata.json |
1 KB |
Model metadata and usage information |
Road Coverage โ 38,874 Named Roads
Every named road in Sri Lanka is indexed โ from major highways (A1 Colombo-Kandy, E01 Southern Expressway) down to local streets (Galle Road, Marine Drive, Ward Place, Kollupitiya Road, etc.). Search by road name, find locations along roads, or discover nearby amenities.
Categories Covered
| Category |
Count |
Examples |
| ๐ฃ๏ธ Named Roads |
38,874 |
A1, E01, Galle Road, Marine Drive |
| ๐ Buildings |
11,751 |
Commercial, residential, government |
| ๐๏ธ Populated Places |
4,482 |
Cities, towns, villages (Colombo, Kandy, Galle, Jaffna, etc.) |
| ๐ช Shops |
4,463 |
Retail stores, supermarkets, pharmacies |
| โช Places of Worship |
4,401 |
Buddhist temples, churches, mosques, kovils |
| ๐ซ Schools |
4,401 |
Government schools, international schools, universities |
| ๐ฅ Hospitals |
902 |
Government hospitals, private clinics, Ayurvedic |
| ๐ Police Stations |
420 |
Police stations and posts |
| ๐ฆ Banks & ATMs |
1,200+ |
Commercial banks (BOC, People's Bank, HNB, Commercial Bank, etc.) |
| ๐จ Hotels |
850+ |
Hotels, guesthouses, resorts |
| ๐ฝ๏ธ Restaurants |
2,000+ |
Restaurants, cafes, food outlets |
| ๐๏ธ Government Offices |
1,500+ |
Divisional secretariats, municipal councils, government departments |
| โฝ Fuel Stations |
600+ |
Petrol stations (Ceypetco, Lanka IOC, etc.) |
| ๐ฟ Natural Features |
2,500+ |
Rivers, mountains, forests, beaches |
| ๐ Transport Hubs |
350+ |
Railway stations, bus stands, airports |
| ๐๏ธ Landmarks |
1,200+ |
Monuments, parks, stadiums, museums |
| ๐ฆ And 10+ more categories |
โ |
Industrial, agricultural, utilities, etc. |
Usage
Quick Start โ Semantic Search
from sentence_transformers import SentenceTransformer
import faiss
import numpy as np
import sqlite3
import json
model = SentenceTransformer('deathlegionteam/sri-lanka-location-intelligence')
index = faiss.read_index('faiss_index_large.bin')
id_mapping = np.load('id_mapping_large.npy', allow_pickle=True)
conn = sqlite3.connect('srilanka_places.db')
cursor = conn.cursor()
def search_locations(query, top_k=10):
"""Search any location in Sri Lanka by semantic query."""
query_vec = model.encode(['query: ' + query], normalize_embeddings=True)
distances, indices = index.search(query_vec.astype(np.float32), top_k)
results = []
for i, idx in enumerate(indices[0]):
location_id = id_mapping[idx]
cursor.execute("SELECT * FROM locations WHERE id=?", (int(location_id),))
row = cursor.fetchone()
if row:
cols = [d[0] for d in cursor.description]
location = dict(zip(cols, row))
location['similarity'] = float(distances[0][i])
results.append(location)
return results
results = search_locations("police station in Colombo")
for r in results[:5]:
print(f"{r['name']} โ {r['category']} ({r['latitude']}, {r['longitude']})")
results = search_locations("Buddhist temple near Kandy")
results = search_locations("hospital with emergency services")
results = search_locations("Colombo 7 restaurant")
Category Filtering
def search_by_category(query, category, top_k=10):
"""Search within a specific category."""
query_vec = model.encode(['query: ' + query], normalize_embeddings=True)
distances, indices = index.search(query_vec.astype(np.float32), top_k * 3)
results = []
for idx in indices[0]:
location_id = id_mapping[idx]
cursor.execute("SELECT * FROM locations WHERE id=? AND category=?",
(int(location_id), category))
row = cursor.fetchone()
if row:
cols = [d[0] for d in cursor.description]
location = dict(zip(cols, row))
location['similarity'] = float(distances[0][list(indices[0]).index(idx)])
results.append(location)
if len(results) >= top_k:
break
return results
results = search_by_category("government hospital", "hospital", 5)
Nearby Search
import math
def haversine(lat1, lon1, lat2, lon2):
"""Calculate distance in km between two coordinates."""
R = 6371
dlat = math.radians(lat2 - lat1)
dlon = math.radians(lon2 - lon1)
a = math.sin(dlat/2)**2 + math.cos(math.radians(lat1)) * \
math.cos(math.radians(lat2)) * math.sin(dlon/2)**2
return R * 2 * math.atan2(math.sqrt(a), math.sqrt(1-a))
def find_nearby(lat, lon, radius_km=5, limit=50):
"""Find locations near a point."""
lat_delta = radius_km / 111.0
lon_delta = radius_km / (111.0 * abs(math.cos(math.radians(lat))) + 0.001)
cursor.execute("""
SELECT * FROM locations
WHERE latitude BETWEEN ? AND ?
AND longitude BETWEEN ? AND ?
ORDER BY ABS(latitude - ?) + ABS(longitude - ?)
LIMIT ?
""", (lat - lat_delta, lat + lat_delta,
lon - lon_delta, lon + lon_delta,
lat, lon, limit * 2))
rows = cursor.fetchall()
cols = [d[0] for d in cursor.description]
results = []
for row in rows:
loc = dict(zip(cols, row))
dist = haversine(lat, lon, loc['latitude'], loc['longitude'])
if dist <= radius_km:
loc['distance_km'] = round(dist, 2)
results.append(loc)
if len(results) >= limit:
break
return results
nearby = find_nearby(6.9344, 79.8428, radius_km=2)
Search Examples (query โ actual results)
| Query |
Finds |
"police emergency" |
Police stations near you |
"Colombo hospital" |
National Hospital of Sri Lanka, private hospitals in Colombo |
"temple tooth relic" |
Temple of the Tooth, Kandy |
"galle road restaurant" |
Restaurants along Galle Road, Colombo |
"Kandy school" |
Schools in Kandy district |
"fuel station A1" |
Petrol stations along the A1 highway |
"beach hotel" |
Beach resorts and hotels |
"central bank" |
Central Bank of Sri Lanka, Colombo |
"เทเทเทเทเถปเถบ" (Sinhala) |
Buddhist temples |
"เฎเฏเฎตเฎฟเฎฒเฏ" (Tamil) |
Hindu temples/kovils |
Data Sources
- OpenStreetMap (via Geofabrik download:
sri-lanka-latest.osm.pbf, 136 MB)
- License: Open Database License (ODbL) โ data from OpenStreetMap contributors
Coverage
- 117,446 total locations indexed
- 27 classification categories
- 38,874 named roads
- 100+ languages supported for search queries
- Full coverage of all 9 provinces (Western, Central, Southern, Northern, Eastern, North Western, North Central, Uva, Sabaragamuwa)
Technical Architecture
User Query (in any language)
โ
multilingual-e5-large (1024-dim embedding)
โ
FAISS IVFFlat Index (nlist=200, cosine similarity)
โ
SQLite Database โ Location Name + Coordinates + Category + Tags
โ
Structured Results with similarity scores
License
- Model: MIT License (intfloat/multilingual-e5-large)
- Location Data: Open Database License (ODbL) โ ยฉ OpenStreetMap contributors
- System: MIT License
Links
