File size: 22,821 Bytes
23804b3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 |
"""
Knowledge Graph Integration for Cyber-LLM
Real-time threat intelligence and cybersecurity knowledge management
Author: Muzan Sano <[email protected]>
"""
import asyncio
import json
import logging
from datetime import datetime, timedelta
from typing import Dict, List, Any, Optional, Tuple, Union, Set
from dataclasses import dataclass, field
from enum import Enum
import numpy as np
import networkx as nx
from neo4j import GraphDatabase
import requests
import feedparser
from bs4 import BeautifulSoup
from ..utils.logging_system import CyberLLMLogger, CyberLLMError, ErrorCategory
from ..memory.persistent_memory import PersistentMemoryManager
class EntityType(Enum):
"""Knowledge graph entity types"""
VULNERABILITY = "vulnerability"
THREAT_ACTOR = "threat_actor"
MALWARE = "malware"
ATTACK_TECHNIQUE = "attack_technique"
INDICATOR = "indicator"
ASSET = "asset"
ORGANIZATION = "organization"
CAMPAIGN = "campaign"
TOOL = "tool"
MITIGATION = "mitigation"
class RelationType(Enum):
"""Knowledge graph relationship types"""
EXPLOITS = "exploits"
MITIGATES = "mitigates"
TARGETS = "targets"
USES = "uses"
ATTRIBUTED_TO = "attributed_to"
SIMILAR_TO = "similar_to"
PART_OF = "part_of"
DETECTS = "detects"
IMPLEMENTS = "implements"
COMMUNICATES_WITH = "communicates_with"
class ConfidenceLevel(Enum):
"""Confidence levels for knowledge assertions"""
LOW = 0.3
MEDIUM = 0.6
HIGH = 0.8
VERY_HIGH = 0.95
@dataclass
class KnowledgeEntity:
"""Knowledge graph entity"""
entity_id: str
entity_type: EntityType
name: str
# Properties
properties: Dict[str, Any] = field(default_factory=dict)
# Metadata
created_at: datetime = field(default_factory=datetime.now)
updated_at: datetime = field(default_factory=datetime.now)
source: Optional[str] = None
confidence: float = 0.8
# Relationships
relationships: List['KnowledgeRelationship'] = field(default_factory=list)
# Tags and classification
tags: Set[str] = field(default_factory=set)
classification: Optional[str] = None
@dataclass
class KnowledgeRelationship:
"""Knowledge graph relationship"""
relationship_id: str
source_entity: str
target_entity: str
relationship_type: RelationType
# Properties
properties: Dict[str, Any] = field(default_factory=dict)
# Metadata
created_at: datetime = field(default_factory=datetime.now)
confidence: float = 0.8
source: Optional[str] = None
# Temporal aspects
valid_from: Optional[datetime] = None
valid_until: Optional[datetime] = None
@dataclass
class ThreatIntelligenceData:
"""Threat intelligence data structure"""
intel_id: str
title: str
description: str
# Classification
threat_type: str
severity: str
confidence: ConfidenceLevel
# Temporal information
discovered_at: datetime
published_at: Optional[datetime] = None
expires_at: Optional[datetime] = None
# Indicators
indicators: List[Dict[str, Any]] = field(default_factory=list)
# Attribution
threat_actors: List[str] = field(default_factory=list)
campaigns: List[str] = field(default_factory=list)
# Source information
source: str
source_reliability: str
# References
references: List[str] = field(default_factory=list)
# Structured data
mitre_techniques: List[str] = field(default_factory=list)
affected_products: List[str] = field(default_factory=list)
class CyberKnowledgeGraph:
"""Comprehensive cybersecurity knowledge graph"""
def __init__(self,
neo4j_uri: str,
neo4j_user: str,
neo4j_password: str,
memory_manager: PersistentMemoryManager,
logger: Optional[CyberLLMLogger] = None):
self.memory_manager = memory_manager
self.logger = logger or CyberLLMLogger(name="knowledge_graph")
# Graph database connection
self.driver = GraphDatabase.driver(neo4j_uri, auth=(neo4j_user, neo4j_password))
# In-memory graph for fast operations
self.graph = nx.MultiDiGraph()
# Entity and relationship tracking
self.entities = {}
self.relationships = {}
# Intelligence sources
self.threat_intel_sources = {}
self.cve_sources = {}
self.news_sources = {}
# Update tracking
self.last_update = {}
self.update_frequencies = {}
# Initialize knowledge graph
asyncio.create_task(self._initialize_knowledge_graph())
self.logger.info("Cyber Knowledge Graph initialized")
async def _initialize_knowledge_graph(self):
"""Initialize knowledge graph with base data"""
try:
# Create database constraints and indexes
await self._create_database_schema()
# Load base cybersecurity ontology
await self._load_base_ontology()
# Initialize threat intelligence sources
await self._initialize_threat_intel_sources()
# Start periodic updates
asyncio.create_task(self._periodic_updates())
self.logger.info("Knowledge graph initialization completed")
except Exception as e:
self.logger.error("Knowledge graph initialization failed", error=str(e))
async def add_entity(self, entity: KnowledgeEntity) -> bool:
"""Add entity to knowledge graph"""
try:
# Store in Neo4j
with self.driver.session() as session:
query = f"""
CREATE (e:{entity.entity_type.value.title()} {{
entity_id: $entity_id,
name: $name,
properties: $properties,
created_at: $created_at,
updated_at: $updated_at,
source: $source,
confidence: $confidence,
tags: $tags,
classification: $classification
}})
"""
session.run(query, {
"entity_id": entity.entity_id,
"name": entity.name,
"properties": json.dumps(entity.properties),
"created_at": entity.created_at.isoformat(),
"updated_at": entity.updated_at.isoformat(),
"source": entity.source,
"confidence": entity.confidence,
"tags": list(entity.tags),
"classification": entity.classification
})
# Store in memory
self.entities[entity.entity_id] = entity
self.graph.add_node(entity.entity_id, **entity.properties)
self.logger.info("Entity added to knowledge graph",
entity_id=entity.entity_id,
entity_type=entity.entity_type.value)
return True
except Exception as e:
self.logger.error("Failed to add entity", error=str(e))
return False
async def add_relationship(self, relationship: KnowledgeRelationship) -> bool:
"""Add relationship to knowledge graph"""
try:
# Store in Neo4j
with self.driver.session() as session:
query = f"""
MATCH (source {{entity_id: $source_entity}})
MATCH (target {{entity_id: $target_entity}})
CREATE (source)-[r:{relationship.relationship_type.value.upper()} {{
relationship_id: $relationship_id,
properties: $properties,
created_at: $created_at,
confidence: $confidence,
source: $source,
valid_from: $valid_from,
valid_until: $valid_until
}}]->(target)
"""
session.run(query, {
"source_entity": relationship.source_entity,
"target_entity": relationship.target_entity,
"relationship_id": relationship.relationship_id,
"properties": json.dumps(relationship.properties),
"created_at": relationship.created_at.isoformat(),
"confidence": relationship.confidence,
"source": relationship.source,
"valid_from": relationship.valid_from.isoformat() if relationship.valid_from else None,
"valid_until": relationship.valid_until.isoformat() if relationship.valid_until else None
})
# Store in memory
self.relationships[relationship.relationship_id] = relationship
self.graph.add_edge(
relationship.source_entity,
relationship.target_entity,
key=relationship.relationship_id,
relationship_type=relationship.relationship_type.value,
**relationship.properties
)
self.logger.info("Relationship added to knowledge graph",
relationship_id=relationship.relationship_id,
relationship_type=relationship.relationship_type.value)
return True
except Exception as e:
self.logger.error("Failed to add relationship", error=str(e))
return False
async def query_entities(self,
entity_type: Optional[EntityType] = None,
properties: Optional[Dict[str, Any]] = None,
tags: Optional[Set[str]] = None) -> List[KnowledgeEntity]:
"""Query entities from knowledge graph"""
try:
# Build query
conditions = []
params = {}
if entity_type:
label = entity_type.value.title()
else:
label = ""
if properties:
for key, value in properties.items():
conditions.append(f"e.properties CONTAINS $prop_{key}")
params[f"prop_{key}"] = json.dumps({key: value})
if tags:
for i, tag in enumerate(tags):
conditions.append(f"$tag_{i} IN e.tags")
params[f"tag_{i}"] = tag
where_clause = " AND ".join(conditions) if conditions else ""
if where_clause:
where_clause = "WHERE " + where_clause
query = f"""
MATCH (e{':' + label if label else ''})
{where_clause}
RETURN e
"""
# Execute query
with self.driver.session() as session:
result = session.run(query, params)
entities = []
for record in result:
node = record["e"]
entity = KnowledgeEntity(
entity_id=node["entity_id"],
entity_type=EntityType(node.labels),
name=node["name"],
properties=json.loads(node.get("properties", "{}")),
created_at=datetime.fromisoformat(node["created_at"]),
updated_at=datetime.fromisoformat(node["updated_at"]),
source=node.get("source"),
confidence=node.get("confidence", 0.8),
tags=set(node.get("tags", [])),
classification=node.get("classification")
)
entities.append(entity)
return entities
except Exception as e:
self.logger.error("Entity query failed", error=str(e))
return []
async def find_paths(self,
source_entity: str,
target_entity: str,
max_depth: int = 3) -> List[List[str]]:
"""Find paths between entities"""
try:
# Use NetworkX for efficient path finding
if self.graph.has_node(source_entity) and self.graph.has_node(target_entity):
paths = list(nx.all_simple_paths(
self.graph,
source_entity,
target_entity,
cutoff=max_depth
))
return paths
return []
except Exception as e:
self.logger.error("Path finding failed", error=str(e))
return []
async def get_entity_neighbors(self, entity_id: str, relationship_types: Optional[List[RelationType]] = None) -> List[KnowledgeEntity]:
"""Get neighboring entities"""
try:
neighbors = []
if entity_id in self.graph:
for neighbor in self.graph.neighbors(entity_id):
if relationship_types:
# Check if any edge has the required relationship type
edges = self.graph[entity_id][neighbor]
for edge_data in edges.values():
if edge_data.get('relationship_type') in [rt.value for rt in relationship_types]:
if neighbor in self.entities:
neighbors.append(self.entities[neighbor])
break
else:
if neighbor in self.entities:
neighbors.append(self.entities[neighbor])
return neighbors
except Exception as e:
self.logger.error("Failed to get entity neighbors", error=str(e))
return []
class ThreatIntelligenceAggregator:
"""Aggregates threat intelligence from multiple sources"""
def __init__(self,
knowledge_graph: CyberKnowledgeGraph,
logger: Optional[CyberLLMLogger] = None):
self.knowledge_graph = knowledge_graph
self.logger = logger or CyberLLMLogger(name="threat_intel")
# Intelligence sources
self.sources = {
"cve": {
"url": "https://cve.mitre.org/data/downloads/",
"update_frequency": timedelta(hours=6)
},
"mitre_attack": {
"url": "https://attack.mitre.org/",
"update_frequency": timedelta(days=1)
},
"threat_feeds": []
}
# Processing state
self.last_updates = {}
self.processing_queue = asyncio.Queue()
# Start processing worker
asyncio.create_task(self._processing_worker())
self.logger.info("Threat Intelligence Aggregator initialized")
async def aggregate_cve_data(self) -> int:
"""Aggregate CVE data from MITRE"""
try:
self.logger.info("Starting CVE data aggregation")
# Fetch CVE JSON feed
cve_url = "https://cve.mitre.org/data/downloads/allitems.json"
async with aiohttp.ClientSession() as session:
async with session.get(cve_url) as response:
if response.status == 200:
cve_data = await response.json()
else:
raise Exception(f"Failed to fetch CVE data: {response.status}")
processed_count = 0
# Process CVE entries
for cve_item in cve_data.get("CVE_Items", []):
cve_id = cve_item["cve"]["CVE_data_meta"]["ID"]
# Create CVE entity
entity = KnowledgeEntity(
entity_id=cve_id,
entity_type=EntityType.VULNERABILITY,
name=cve_id,
properties={
"description": cve_item["cve"]["description"]["description_data"][0]["value"],
"published_date": cve_item.get("publishedDate"),
"modified_date": cve_item.get("lastModifiedDate"),
"cvss_score": self._extract_cvss_score(cve_item),
"severity": self._determine_severity(cve_item),
"affected_products": self._extract_affected_products(cve_item)
},
source="mitre_cve",
confidence=0.95
)
await self.knowledge_graph.add_entity(entity)
processed_count += 1
# Add relationships to affected products
for product in entity.properties.get("affected_products", []):
# Create or get product entity
product_entity = await self._get_or_create_product_entity(product)
# Create vulnerability relationship
relationship = KnowledgeRelationship(
relationship_id=f"{cve_id}_affects_{product_entity.entity_id}",
source_entity=cve_id,
target_entity=product_entity.entity_id,
relationship_type=RelationType.TARGETS,
confidence=0.9,
source="mitre_cve"
)
await self.knowledge_graph.add_relationship(relationship)
self.last_updates["cve"] = datetime.now()
self.logger.info("CVE data aggregation completed",
processed_count=processed_count)
return processed_count
except Exception as e:
self.logger.error("CVE data aggregation failed", error=str(e))
return 0
async def aggregate_mitre_attack(self) -> int:
"""Aggregate MITRE ATT&CK framework data"""
try:
self.logger.info("Starting MITRE ATT&CK data aggregation")
# MITRE ATT&CK STIX data
attack_url = "https://raw.githubusercontent.com/mitre/cti/master/enterprise-attack/enterprise-attack.json"
async with aiohttp.ClientSession() as session:
async with session.get(attack_url) as response:
if response.status == 200:
attack_data = await response.json()
else:
raise Exception(f"Failed to fetch MITRE ATT&CK data: {response.status}")
processed_count = 0
# Process STIX objects
for stix_object in attack_data.get("objects", []):
if stix_object["type"] == "attack-pattern":
# Create technique entity
technique_id = stix_object.get("external_references", [{}])[0].get("external_id", "")
entity = KnowledgeEntity(
entity_id=technique_id,
entity_type=EntityType.ATTACK_TECHNIQUE,
name=stix_object["name"],
properties={
"description": stix_object.get("description", ""),
"kill_chain_phases": [phase["phase_name"] for phase in stix_object.get("kill_chain_phases", [])],
"platforms": stix_object.get("x_mitre_platforms", []),
"tactics": [ref["external_id"] for ref in stix_object.get("external_references", []) if ref.get("source_name") == "mitre-attack"]
},
source="mitre_attack",
confidence=0.98
)
await self.knowledge_graph.add_entity(entity)
processed_count += 1
self.last_updates["mitre_attack"] = datetime.now()
self.logger.info("MITRE ATT&CK data aggregation completed",
processed_count=processed_count)
return processed_count
except Exception as e:
self.logger.error("MITRE ATT&CK data aggregation failed", error=str(e))
return 0
async def _processing_worker(self):
"""Background worker for processing intelligence data"""
while True:
try:
# Check for scheduled updates
for source, config in self.sources.items():
last_update = self.last_updates.get(source)
update_frequency = config.get("update_frequency")
if not last_update or (datetime.now() - last_update) > update_frequency:
if source == "cve":
await self.aggregate_cve_data()
elif source == "mitre_attack":
await self.aggregate_mitre_attack()
# Wait before next check
await asyncio.sleep(3600) # Check every hour
except Exception as e:
self.logger.error("Intelligence processing worker error", error=str(e))
await asyncio.sleep(300) # Wait 5 minutes on error
# Factory functions
def create_knowledge_graph(neo4j_uri: str,
neo4j_user: str,
neo4j_password: str,
memory_manager: PersistentMemoryManager,
**kwargs) -> CyberKnowledgeGraph:
"""Create cyber knowledge graph"""
return CyberKnowledgeGraph(neo4j_uri, neo4j_user, neo4j_password, memory_manager, **kwargs)
def create_threat_intelligence_aggregator(knowledge_graph: CyberKnowledgeGraph,
**kwargs) -> ThreatIntelligenceAggregator:
"""Create threat intelligence aggregator"""
return ThreatIntelligenceAggregator(knowledge_graph, **kwargs)
|