--- license: mit task_categories: - tabular-classification tags: - cybersecurity - malware-detection - phishing-detection - infrastructure-metadata - domains configs: - config_name: domainradar-malware data_files: - split: train path: DataInBriefHranicky/malware/train.csv - split: validation path: DataInBriefHranicky/malware/dev.csv - split: test path: DataInBriefHranicky/malware/test.csv - config_name: domainradar-phishing data_files: - split: train path: DataInBriefHranicky/phishing/train.csv - split: validation path: DataInBriefHranicky/phishing/dev.csv - split: test path: DataInBriefHranicky/phishing/test.csv - config_name: domainradar-malware-family data_files: - split: train path: DataInBriefHranicky/malware_classes/train.csv - split: validation path: DataInBriefHranicky/malware_classes/dev.csv - split: test path: DataInBriefHranicky/malware_classes/test.csv - config_name: threatfox-malware data_files: - split: train path: ThreatFox/malware/train.csv - split: validation path: ThreatFox/malware/dev.csv - split: test path: ThreatFox/malware/test.csv - config_name: threatfox-malware-family data_files: - split: train path: ThreatFox/malware_types/train.csv - split: validation path: ThreatFox/malware_types/dev.csv - split: test path: ThreatFox/malware_types/test.csv - config_name: deepurlbench-malware data_files: - split: train path: DeepURLBench/malware/train.csv - split: validation path: DeepURLBench/malware/dev.csv - split: test path: DeepURLBench/malware/test.csv - config_name: deepurlbench-phishing data_files: - split: train path: DeepURLBench/phishing/train.csv - split: validation path: DeepURLBench/phishing/dev.csv - split: test path: DeepURLBench/phishing/test.csv - config_name: macformer-malware data_files: - split: train path: MACFormer/malware/train.csv - split: validation path: MACFormer/malware/dev.csv - split: test path: MACFormer/malware/test.csv - config_name: macformer-phishing data_files: - split: train path: MACFormer/phishing/train.csv - split: validation path: MACFormer/phishing/dev.csv - split: test path: MACFormer/phishing/test.csv pretty_name: MIAF Domain Detection Infrastructure Datasets size_categories: - 100K This collection is the standardized evaluation benchmark for **MIAF (Modular Infrastructure-Aware Fusion)**. It provides nine classification datasets derived from four public malicious-domain benchmarks, each paired with a shared 137-feature infrastructure representation. ## Overview We evaluate MIAF across nine classification datasets derived from four public malicious-domain benchmarks: - **[DomainRadar](https://doi.org/10.1016/j.dib.2025.112062)** (Hranický et al., 2025) - **[ThreatFox](https://threatfox.abuse.ch/)** (abuse.ch) - **[DeepURLBench](https://arxiv.org/abs/2501.00356)** (Schvartzman et al., 2024) - **[MACFormer/CURL-IP](https://doi.org/10.1109/TIFS.2026.3709128)** (Tian et al., 2026) These datasets cover malware detection, phishing detection, and long-tailed malware-family attribution. The benchmarks span diverse characteristics in terms of scale, class balance, label granularity, and metadata availability, enabling an evaluation of the efficacy of infrastructure metadata under varied operating conditions. ## Quick Start Install the [Datasets](https://huggingface.co/docs/datasets/) library and load one of the nine task-specific configurations: ```python from datasets import load_dataset dataset = load_dataset( "ahmedBargady/MIAF_DomainDetection_Infrastructure_Datasets", "threatfox-malware", ) train = dataset["train"] validation = dataset["validation"] test = dataset["test"] ``` Available configurations: `domainradar-malware`, `domainradar-phishing`, `domainradar-malware-family`, `threatfox-malware`, `threatfox-malware-family`, `deepurlbench-malware`, `deepurlbench-phishing`, `macformer-malware`, and `macformer-phishing`. ## Data Splits and Reorganization To establish a consistent evaluation baseline, we have reorganized each benchmark into task-specific subsets and partitioned them into disjoint **training**, **development**, and **testing** sets using a **60/20/20** ratio. This stratification was applied to every individual class within the datasets to ensure a fair and consistent baseline across all evaluated methods. All datasets that lacked native infrastructure metadata (ThreatFox, DeepURLBench, and MACFormer) have been enriched using the uniform EDC pipeline. ## Infrastructure Metadata Each row in `train.csv`, `dev.csv`, and `test.csv` pairs a domain name with **137 standardized numerical infrastructure features**. DomainRadar (stored under `DataInBriefHranicky/`) provides native EDC metadata; we collected the same metadata for ThreatFox, DeepURLBench, and MACFormer with the Enrichment Data Collector (EDC), giving all nine task datasets one shared feature schema. The 137 fields comprise five semantic metadata categories plus two input-validation flags: | Group | Features | What it captures | | :--- | ---: | :--- | | **Input validation** | 2 | Whether the input parsed successfully and whether it is a literal IP address. | | **RDAP/WHOIS** | 37 | Registration age, lifecycle, privacy, nameserver, status, and entity properties. | | **DNS** | 29 | Resolution, record counts, TTLs, DNS providers, mail, DNSSEC, CNAME, and SOA properties. | | **TLS** | 26 | Certificate age, validity, issuer, SAN, wildcard, and expiration properties. | | **IP/Hosting** | 31 | Resolved-address types, ASN/country/RIR diversity, network age, and hosting-provider properties. | | **Cross-signal** | 12 | Consistency and interaction features computed across registration, DNS, TLS, and hosting observations. | Continuous quantities (for example, registration age and TTL) remain numerical. Boolean properties are binary indicators. High-cardinality concepts such as ASNs, nameservers, and hosting networks are represented through security-specific indicators and diversity counts rather than raw identifiers. Missing infrastructure observations are retained as empty CSV cells. Presence and availability fields preserve observable missingness where available; the MIAF experiments set remaining missing values to zero during data loading. Infrastructure metadata changes over time, so these values describe the collection-time snapshot and may not match the infrastructure present when a domain was originally labeled. ### Complete Feature Catalog
Input validation (2 features) | Feature | Description | | :--- | :--- | | `input_valid` | Input parsed to a valid domain/hostname | | `input_is_ip_address` | Input is a literal IP address, not a domain |
RDAP/WHOIS (37 features) | Feature | Description | | :--- | :--- | | `rdap_domain_age_days` | Days since domain registration (vs scan date) | | `rdap_registration_period_days` | Days between registration and expiration | | `rdap_days_since_update` | Days since the registration record last changed | | `rdap_days_to_expiration` | Days until registration expires | | `rdap_is_new_domain_1_day` | Registered within the last 1 day | | `rdap_is_new_domain_7_days` | Registered within the last 7 days | | `rdap_is_new_domain_30_days` | Registered within the last 30 days | | `rdap_is_new_domain_90_days` | Registered within the last 90 days | | `rdap_is_new_domain_365_days` | Registered within the last 365 days | | `rdap_registered_for_one_year` | Registration period ≤ ~1 year | | `rdap_registered_for_two_years_or_less` | Registration period ≤ ~2 years | | `rdap_registered_for_five_years_or_more` | Registration period ≥ ~5 years | | `rdap_expires_soon_7_days` | Expires within 7 days | | `rdap_expires_soon_30_days` | Expires within 30 days | | `rdap_expires_soon_90_days` | Expires within 90 days | | `rdap_recently_updated_7_days` | Record updated within last 7 days | | `rdap_recently_updated_30_days` | Record updated within last 30 days | | `rdap_recently_updated_90_days` | Record updated within last 90 days | | `rdap_privacy_protected` | WHOIS/RDAP privacy or redaction terms present | | `rdap_num_nameservers` | Number of nameservers listed | | `rdap_num_unique_nameserver_domains` | Number of distinct nameserver base domains | | `rdap_uses_cloudflare_ns` | Uses Cloudflare nameservers | | `rdap_uses_google_ns` | Uses Google nameservers | | `rdap_uses_amazon_ns` | Uses Amazon/AWS nameservers | | `rdap_uses_azure_ns` | Uses Azure/Microsoft nameservers | | `rdap_uses_digitalocean_ns` | Uses DigitalOcean nameservers | | `rdap_uses_registrar_default_ns` | Uses registrar default/parking nameservers | | `rdap_num_status_codes` | Number of EPP/RDAP status codes | | `rdap_has_client_transfer_prohibited` | clientTransferProhibited status set | | `rdap_has_server_transfer_prohibited` | serverTransferProhibited status set | | `rdap_has_client_hold` | clientHold status set | | `rdap_has_server_hold` | serverHold status set | | `rdap_has_pending_delete` | pendingDelete status set | | `rdap_has_redemption_period` | redemptionPeriod status set | | `rdap_has_any_hold_status` | Any hold status present | | `rdap_has_any_delete_status` | Any delete/redemption status present | | `rdap_entity_count` | Number of related entities (registrar, registrant, ...) |
DNS (29 features) | Feature | Description | | :--- | :--- | | `dns_resolves` | Domain resolves to at least one IP | | `dns_num_a_records` | Number of A (IPv4) records | | `dns_num_aaaa_records` | Number of AAAA (IPv6) records | | `dns_num_unique_ips` | Number of distinct resolved IPs | | `dns_min_ttl` | Minimum TTL across records | | `dns_max_ttl` | Maximum TTL across records | | `dns_mean_ttl` | Mean TTL across records | | `dns_has_low_ttl_60` | Minimum TTL ≤ 60s | | `dns_has_low_ttl_300` | Minimum TTL ≤ 300s | | `dns_num_ns_records` | Number of NS records | | `dns_num_unique_ns_domains` | Number of distinct NS base domains | | `dns_uses_cloudflare_dns` | NS hosted by Cloudflare | | `dns_uses_google_dns` | NS hosted by Google | | `dns_uses_amazon_dns` | NS hosted by Amazon/AWS | | `dns_uses_azure_dns` | NS hosted by Azure/Microsoft | | `dns_uses_godaddy_dns` | NS hosted by GoDaddy | | `dns_uses_namecheap_dns` | NS hosted by Namecheap | | `dns_num_mx_records` | Number of MX (mail) records | | `dns_uses_google_mail` | MX points to Google mail | | `dns_uses_microsoft_mail` | MX points to Microsoft/Outlook mail | | `dns_uses_protonmail` | MX points to ProtonMail | | `dns_uses_zoho_mail` | MX points to Zoho mail | | `dns_num_txt_records` | Number of TXT records | | `dns_has_spf_record` | SPF record present (v=spf1) | | `dns_num_cname_records` | CNAME record present (0/1) | | `dns_has_soa_record` | SOA record present | | `dns_has_dnssec` | DNSSEC present (has DNSKEY) | | `dns_cname_ends_in_free_hoster` | CNAME target is a free hosting provider | | `dns_cname_ends_in_cdn` | CNAME target is a CDN |
TLS (26 features) | Feature | Description | | :--- | :--- | | `tls_certificate_age_days` | Days since certificate validity start | | `tls_validity_period_days` | Certificate validity window length (days) | | `tls_days_to_expiration` | Days until certificate expires | | `tls_certificate_expired` | Certificate already expired | | `tls_certificate_not_yet_valid` | Certificate not yet valid | | `tls_issuer_is_lets_encrypt` | Issued by Let's Encrypt | | `tls_issuer_is_google` | Issued by Google Trust Services | | `tls_issuer_is_cloudflare` | Issued by Cloudflare | | `tls_issuer_is_digicert` | Issued by DigiCert | | `tls_issuer_is_sectigo` | Issued by Sectigo/Comodo | | `tls_issuer_is_zerossl` | Issued by ZeroSSL | | `tls_issuer_is_amazon` | Issued by Amazon | | `tls_num_san_domains` | Number of Subject Alternative Names | | `tls_has_wildcard_san` | Any wildcard (*.) SAN present | | `tls_has_many_sans_10` | ≥10 SANs | | `tls_has_many_sans_50` | ≥50 SANs | | `tls_any_san_matches_host` | A SAN matches the queried host | | `tls_new_cert_1_day` | Certificate issued within last 1 day | | `tls_new_cert_7_days` | Certificate issued within last 7 days | | `tls_new_cert_30_days` | Certificate issued within last 30 days | | `tls_expires_soon_7_days` | Certificate expires within 7 days | | `tls_expires_soon_30_days` | Certificate expires within 30 days | | `tls_expires_soon_90_days` | Certificate expires within 90 days | | `tls_short_validity_90_days` | Validity window ≤ 90 days | | `tls_short_validity_180_days` | Validity window ≤ 180 days | | `tls_long_validity_398_days` | Validity window > 398 days |
IP/Hosting (31 features) | Feature | Description | | :--- | :--- | | `ip_num_resolved_ips` | Number of resolved IP addresses | | `ip_num_ipv4` | Number of IPv4 addresses | | `ip_num_ipv6` | Number of IPv6 addresses | | `ip_has_public_ip` | Has a public/global IP | | `ip_has_private_ip` | Has a private IP | | `ip_has_reserved_ip` | Has a reserved IP | | `ip_has_loopback_ip` | Has a loopback IP | | `ip_has_multicast_ip` | Has a multicast IP | | `ip_has_link_local_ip` | Has a link-local IP | | `ip_num_unique_asns` | Number of distinct ASNs | | `ip_num_unique_countries` | Number of distinct hosting countries | | `ip_num_unique_registries` | Number of distinct RIRs (ARIN/RIPE/...) | | `ip_num_unique_networks` | Number of distinct network names/orgs | | `ip_network_registration_age_days_min` | Min IP-network allocation age (days) | | `ip_network_registration_age_days_mean` | Mean IP-network allocation age (days) | | `ip_network_recently_allocated_30d` | A hosting network allocated within 30 days | | `ip_network_recently_allocated_90d` | A hosting network allocated within 90 days | | `ip_uses_cloud_hosting` | Hosted on a major cloud provider | | `ip_uses_cdn_or_proxy` | Behind a CDN or proxy | | `ip_uses_amazon_hosting` | Hosted on Amazon/AWS | | `ip_uses_google_hosting` | Hosted on Google | | `ip_uses_microsoft_hosting` | Hosted on Microsoft/Azure | | `ip_uses_cloudflare_hosting` | Hosted on/behind Cloudflare | | `ip_uses_digitalocean_hosting` | Hosted on DigitalOcean | | `ip_uses_ovh_hosting` | Hosted on OVH | | `ip_uses_hetzner_hosting` | Hosted on Hetzner | | `ip_uses_godaddy_hosting` | Hosted on GoDaddy | | `ip_uses_namecheap_hosting` | Hosted on Namecheap | | `ip_multi_asn_hosting` | IPs span more than one ASN | | `ip_multi_country_hosting` | IPs span more than one country | | `ip_multi_registry_hosting` | IPs span more than one RIR |
Cross-signal (12 features) | Feature | Description | | :--- | :--- | | `x_age_cert_vs_domain_ratio` | Cert age ÷ (domain age + 1) | | `x_ttl_to_domain_age_ratio` | Min TTL ÷ (domain age + 1) | | `x_domain_and_cert_both_new_7d` | Domain and cert both ≤7 days old | | `x_domain_new_but_no_dns` | New domain (≤30d) that doesn't resolve | | `x_low_ttl_on_old_domain` | Low TTL on a domain older than 1 year | | `x_dns_ip_count_vs_hosting_ip_count_diff` | \|DNS unique IPs - resolved hosting IPs\| | | `x_rdap_dns_ns_count_diff` | \|RDAP NS count - DNS NS count\| | | `x_rdap_dns_ns_mismatch` | RDAP vs DNS NS count differ by ≥2 | | `x_ns_provider_consistency_cloudflare` | RDAP and DNS agree on Cloudflare NS | | `x_ns_provider_consistency_google` | RDAP and DNS agree on Google NS | | `x_ns_provider_consistency_amazon` | RDAP and DNS agree on Amazon NS | | `x_dnssec_and_old_domain` | DNSSEC enabled on a domain older than 1 year |
## Dataset Characteristics | Dataset | Task | Configuration | Samples | Classes | Malicious share (each split) | Metadata source | | :--- | :--- | :--- | ---: | :--- | ---: | :--- | | **DomainRadar** | Malware (binary) | `domainradar-malware` | 341,935 | 2 (legit / malicious) | 19.1% | Native | | | Phishing (binary) | `domainradar-phishing` | 441,207 | 2 (legit / malicious) | 37.3% | Native | | | Malware family (multiclass) | `domainradar-malware-family` | 311,906 | 61 (legit + 60 families) | 11.3% | Native | | **ThreatFox** | Malware (binary) | `threatfox-malware` | 176,065 | 2 (legit / malicious) | 43.2% | EDC-enriched | | | Malware family (multiclass) | `threatfox-malware-family` | 176,065 | 65 (legit + 64 families) | 43.2% | EDC-enriched | | **DeepURLBench** | Malware (binary) | `deepurlbench-malware` | 375,917 | 2 (benign / malicious) | 24.1% | EDC-enriched | | | Phishing (binary) | `deepurlbench-phishing` | 362,347 | 2 (benign / phishing) | 21.3% | EDC-enriched | | **MACFormer** | Malware (binary) | `macformer-malware` | 350,532 | 2 (legit / malicious) | 32.9% | EDC-enriched | | | Phishing (binary) | `macformer-phishing` | 266,500 | 2 (legit / malicious) | 11.8% | EDC-enriched | ## Structure Inside each task folder: - `train.csv`: Training split (60%) with domain names, labels, and the 137-feature infrastructure vector. - `dev.csv`: Development/validation split (20%) with the same schema. - `test.csv`: Testing split (20%) with the same schema. - `*_ip.csv`: Resolved-IP sidecar records corresponding to each split; sidecar columns vary by source benchmark. ## Data Fields | Field | Description | | :--- | :--- | | `input` | Domain name or hostname. | | `class` | Task-specific class; use this as the target for malware-family attribution. | | `label` | Coarse binary label; use this as the target for malware and phishing detection. | | `label_int` | Integer label encoding included in DomainRadar-derived files. It is a target field, not an infrastructure feature. | | `_error`, `_elapsed_s` | Collection diagnostics present in some EDC-enriched files. They are not among the 137 MIAF features and should not be used as model inputs when reproducing the benchmark. | | 137 infrastructure columns | The input-validation, RDAP/WHOIS, DNS, TLS, IP/hosting, and cross-signal fields cataloged above. | For MIAF-compatible metadata input, exclude `input`, `class`, `label`, `label_int`, `_error`, and `_elapsed_s` when those columns are present. Preserve the feature columns in their CSV order. ## Intended Uses This release supports research and reproducible benchmarking in: - infrastructure-aware malicious domain-name and phishing detection; - long-tailed malware-family attribution; - multimodal fusion of domain-name strings and infrastructure metadata; - metadata-category ablation, missing-data robustness, and temporal-drift studies; and - interpretable analysis of registration, DNS, TLS, and hosting evidence. The data is intended for research and evaluation. A model trained on these snapshots should not be used as the sole basis for production blocking or attribution decisions. ## Limitations and Responsible Use - Infrastructure metadata is time-dependent. The released values describe the collection-time snapshot and may not match the infrastructure present when a domain was first observed or labeled. - Missing values are expected because DNS, RDAP/WHOIS, TLS, and IP/hosting sources differ in coverage, availability, latency, and rate limits. - Labels inherit the scope, collection procedures, class balance, and possible noise of the four source benchmarks. - Domain names and resolved IP records may identify active or formerly active malicious infrastructure. Do not visit or execute content from listed domains without an isolated security-analysis environment. - The standardized feature vectors contain no raw WHOIS/RDAP contact names, email addresses, or postal addresses. ## Citation and Attribution If you use this collection, cite the dataset and the relevant source benchmark linked in the [Overview](#overview): ```bibtex @dataset{bargady_2026_miaf_infrastructure, author = {Ahmed Bargady}, title = {MIAF: Domain Detection Infrastructure Datasets}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/datasets/ahmedBargady/MIAF_DomainDetection_Infrastructure_Datasets} } ``` ## License This standardized release is provided under the MIT license declared in the dataset card metadata. The underlying benchmarks remain attributable to their original creators; review their linked publications and terms when redistributing derived data.