{
    "data_id": "43622",
    "name": "Binary-Dataset-of-Phishing-and-Legitimate-URLs",
    "exact_name": "Binary-Dataset-of-Phishing-and-Legitimate-URLs",
    "version": 1,
    "version_label": "v1.0",
    "description": "Description\nThe data set is provided csv file which provides the following resources that can be used as inputs for model building :\nA collection of website URLs for 11001 websites. Each sample has 15 website parameters and a class label identifying it as a phishing website or not (0 or 1).\nIf URLs is Phished then label is 0 and for legitimate label is 1\nThe data set also serves as an input for project scoping and tries to specify the functional and non-functional requirements for it.",
    "format": "arff",
    "uploader": "Dustin Carrion",
    "uploader_id": 30123,
    "visibility": "public",
    "creator": null,
    "contributor": null,
    "date": "2022-03-24 00:42:33",
    "update_comment": null,
    "last_update": "2022-03-24 00:42:33",
    "licence": "Attribution 4.0 International (CC BY 4.0)",
    "status": "active",
    "error_message": null,
    "url": "https:\/\/www.openml.org\/data\/download\/22102447\/dataset",
    "default_target_attribute": null,
    "row_id_attribute": null,
    "ignore_attribute": null,
    "runs": 0,
    "suggest": {
        "input": [
            "Binary-Dataset-of-Phishing-and-Legitimate-URLs",
            "Description The data set is provided csv file which provides the following resources that can be used as inputs for model building : A collection of website URLs for 11001 websites. Each sample has 15 website parameters and a class label identifying it as a phishing website or not (0 or 1). If URLs is Phished then label is 0 and for legitimate label is 1 The data set also serves as an input for project scoping and tries to specify the functional and non-functional requirements for it. "
        ],
        "weight": 5
    },
    "qualities": {
        "NumberOfInstances": 11000,
        "NumberOfFeatures": 15,
        "NumberOfClasses": null,
        "NumberOfMissingValues": 0,
        "NumberOfInstancesWithMissingValues": 0,
        "NumberOfNumericFeatures": 15,
        "NumberOfSymbolicFeatures": 0,
        "Dimensionality": 0.0013636363636363637,
        "PercentageOfNumericFeatures": 100,
        "MajorityClassPercentage": null,
        "PercentageOfSymbolicFeatures": 0,
        "MajorityClassSize": null,
        "MinorityClassPercentage": null,
        "MinorityClassSize": null,
        "NumberOfBinaryFeatures": 0,
        "PercentageOfBinaryFeatures": 0,
        "PercentageOfInstancesWithMissingValues": 0,
        "AutoCorrelation": null,
        "PercentageOfMissingValues": 0
    },
    "tags": [
        {
            "uploader": "38960",
            "tag": "Computer Systems"
        },
        {
            "uploader": "38960",
            "tag": "Machine Learning"
        }
    ],
    "features": [
        {
            "name": "whois_regDate",
            "index": "0",
            "type": "numeric",
            "distinct": "4038",
            "missing": "0",
            "min": "-1",
            "max": "12915",
            "mean": "3051",
            "stdev": "3206"
        },
        {
            "name": "whois_expDate",
            "index": "1",
            "type": "numeric",
            "distinct": "1872",
            "missing": "0",
            "min": "-327",
            "max": "31868",
            "mean": "463",
            "stdev": "756"
        },
        {
            "name": "whois_updatedDate",
            "index": "2",
            "type": "numeric",
            "distinct": "1146",
            "missing": "0",
            "min": "-1",
            "max": "3685",
            "mean": "251",
            "stdev": "391"
        },
        {
            "name": "dot_count",
            "index": "3",
            "type": "numeric",
            "distinct": "19",
            "missing": "0",
            "min": "1",
            "max": "28",
            "mean": "2",
            "stdev": "2"
        },
        {
            "name": "url_len",
            "index": "4",
            "type": "numeric",
            "distinct": "347",
            "missing": "0",
            "min": "4",
            "max": "2123",
            "mean": "49",
            "stdev": "77"
        },
        {
            "name": "digit_count",
            "index": "5",
            "type": "numeric",
            "distinct": "128",
            "missing": "0",
            "min": "0",
            "max": "702",
            "mean": "6",
            "stdev": "18"
        },
        {
            "name": "special_count",
            "index": "6",
            "type": "numeric",
            "distinct": "35",
            "missing": "0",
            "min": "0",
            "max": "59",
            "mean": "1",
            "stdev": "4"
        },
        {
            "name": "hyphen_count",
            "index": "7",
            "type": "numeric",
            "distinct": "32",
            "missing": "0",
            "min": "0",
            "max": "32",
            "mean": "1",
            "stdev": "3"
        },
        {
            "name": "double_slash",
            "index": "8",
            "type": "numeric",
            "distinct": "6",
            "missing": "0",
            "min": "0",
            "max": "6",
            "mean": "1",
            "stdev": "1"
        },
        {
            "name": "single_slash",
            "index": "9",
            "type": "numeric",
            "distinct": "25",
            "missing": "0",
            "min": "0",
            "max": "25",
            "mean": "2",
            "stdev": "3"
        },
        {
            "name": "at_the_rate",
            "index": "10",
            "type": "numeric",
            "distinct": "4",
            "missing": "0",
            "min": "0",
            "max": "3",
            "mean": "0",
            "stdev": "0"
        },
        {
            "name": "protocol",
            "index": "11",
            "type": "numeric",
            "distinct": "2",
            "missing": "0",
            "min": "0",
            "max": "1",
            "mean": "0",
            "stdev": "0"
        },
        {
            "name": "protocol_count",
            "index": "12",
            "type": "numeric",
            "distinct": "6",
            "missing": "0",
            "min": "0",
            "max": "6",
            "mean": "1",
            "stdev": "1"
        },
        {
            "name": "web_traffic",
            "index": "13",
            "type": "numeric",
            "distinct": "2",
            "missing": "0",
            "min": "0",
            "max": "1",
            "mean": "0",
            "stdev": "0"
        },
        {
            "name": "label",
            "index": "14",
            "type": "numeric",
            "distinct": "2",
            "missing": "0",
            "min": "0",
            "max": "1",
            "mean": "1",
            "stdev": "1"
        }
    ],
    "nr_of_issues": 0,
    "nr_of_downvotes": 0,
    "nr_of_likes": 0,
    "nr_of_downloads": 0,
    "total_downloads": 0,
    "reach": 0,
    "reuse": 0,
    "impact_of_reuse": 0,
    "reach_of_reuse": 0,
    "impact": 0
}