Dataset Viewer

The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

talby/spamassassin(複製)

データの内容

SpamAssassin 公開メールコーパスのデータで、スパムフィルタの試験に使うためにプロジェクトのメンバーが選んだメールの集合である。1 行が 1 通のメールで、label に spam か ham、group に 5 種類のカテゴリ、text に本文が入る。text 設定は文字セットを utf8 に正規化し、MIME ツリーを JSON のリストのリストとして書き出している。提供されているのは train 分割だけである。

データの概要

項目 値
行数 21,500(2 構成・2 分割の合計)
列数 3
容量 564.30 KB(ファイル 4 本)
ファイルの形式 py 2
公開 public

構成と分割

構成(config) 分割(split)ごとの行数 列数
text train 10,749 3
unprocessed train 10,751 3

列(構成 text・分割 train。構成ごとに列が違う)

列 型 先頭の行の値
label class_label(2: spam, ham) 1
group string easy_ham
text string "\n> From: fork-admin@xent.com [mailto:fork-admi…

行数と列は HF のデータセットビューアの集計(元のデータセットの現在の版で数えたもの)。全データセットの一覧は HYUKI5/dataset-catalog にある。

複製について

talby/spamassassin をリビジョン a1b3371d2df15575f766c081cb329543955d27f6 の時点でそのまま複製したもの。中身は変えていない。判断モデル Laya(convaiinnovations/laya)の微調整に使う候補として、調査メモ(Laya 学習用データセット調査)に載せたデータを集めている。

出典

論文

  • 元のカードに論文へのリンクは見つからなかった

アノテーションの方法

収集の理由

このデータセットによって、現代の NLP ツールが古い NLP の問題を解けることを検証できることが期待されている。

最初のデータ収集と正規化

上流のコーパスの説明 に、収集方法の詳細が書かれている。ここでテキスト本文を復元する作業は、主に email.parser と ftfy で行われている。

微調整に使うときの注意

  • 調査メモの台帳での区分: D(記載なし・other・unknown・資料間の食い違い)。区分は記載の文言で機械的に付けたもので、法的な判定ではない。学習に入れる前に元のカードとライセンスの原文で確かめる。
  • Laya の入力は 2,048 token まで(ModernBERT-large / mmBERT)。長い本文は先頭から黙って切られるので、切れる行は事前に除くか、要約・切り出しをしてから入れる。
  • 分割は記事・文書単位で行い、同じ本文から作った別の問いを学習側と較正側に分けない。温度と閾値は学習に使っていない分割で別々に当てる。

調査メモの記述

メモ 節 ライセンス(記載) 確認 注記
routing_support_structured.md Q3. サポート / チケット / メール / レビュー系 (L97) unknown 確認済 確認印なし(各ファイルの凡例により確認済扱い) / 原文注記: HF API 確認

元のカード(日本語訳)

以下は元のデータセットのカードの本文を日本語に訳したもの(機械翻訳)。訳に疑問があるときは、末尾の原文を確かめる。


SpamAssassin 公開メールコーパスのデータセットカード

データセットの説明

データセットの概要

スパムフィルタリングシステムの試験に使うために、SpamAssassin プロジェクトのメンバーが選んだメールの集合である。

対応するタスクとリーダーボード

[More Information Needed]

言語

[More Information Needed]

データセットの構造

データの例

  • text 設定は、すべての文字セットを utf8 に正規化し、 MIME ツリーを JSON のリストのリストとして書き出す。
  • unprocessed 設定は、メッセージを一切解析せず、 ヘッダーと内容の全体をバイナリのまま残す。

データのフィールド

  • label: spam または ham
  • group: SpamAssassin はこれらのサンプルをカテゴリに分類している {'hard_ham', 'spam_2', 'spam', 'easy_ham', 'easy_ham_2'}
  • text: メッセージ本文を正規化したテキスト
  • raw: メッセージの完全なバイナリのヘッダーと内容

データの分割

提供されているのは train 分割のみである。

データセットの作成

収集の理由

このデータセットによって、現代の NLP ツールが古い NLP の問題を解けることを検証できることが期待されている。

元データ

最初のデータ収集と正規化

上流のコーパスの説明 に、収集方法の詳細が書かれている。ここでテキスト本文を復元する作業は、主に email.parser と ftfy で行われている。

元の言語の作成者は誰か?

[More Information Needed]

アノテーション

アノテーションの過程

[More Information Needed]

アノテーターは誰か?

[More Information Needed]

個人情報と機微な情報

[More Information Needed]

データ利用上の留意点

データセットの社会的影響

[More Information Needed]

バイアスに関する検討

[More Information Needed]

その他の既知の制限

[More Information Needed]

追加情報

データセットの作成者

[More Information Needed]

ライセンス情報

[More Information Needed]

引用情報

[More Information Needed]

貢献者

[More Information Needed]

原文(英語)

Dataset Card for the SpamAssassin public mail corpus

Dataset Summary

This is a selection of mail messages, suitable for use in testing spam filtering systems assembled by members of the SpamAssassin project.

Supported Tasks and Leaderboards

[More Information Needed]

Languages

[More Information Needed]

Dataset Structure

Data Instances

  • The text config normalizes all character sets to utf8 and dumps the MIME tree as a JSON list of lists.
  • The unprocessed config does not parse messages at all, leaving the full headers and content as binary.

Data Fields

  • label: spam or ham
  • group: SpamAssassin has grouped these samples into categories {'hard_ham', 'spam_2', 'spam', 'easy_ham', 'easy_ham_2'}
  • text: normalized text of the message bodies
  • raw: full binary headers and contents of messages

Data Splits

Only a train split has been provided.

Dataset Creation

Curation Rationale

It is hoped this dataset can help verify that modern NLP tools can solve old NLP problems.

Source Data

Initial Data Collection and Normalization

The upstream corpus description goes into detail on collection methods. The work here to recover text bodies is largely done with email.parser and ftfy.

Who are the source language producers?

[More Information Needed]

Annotations

Annotation process

[More Information Needed]

Who are the annotators?

[More Information Needed]

Personal and Sensitive Information

[More Information Needed]

Considerations for Using the Data

Social Impact of Dataset

[More Information Needed]

Discussion of Biases

[More Information Needed]

Other Known Limitations

[More Information Needed]

Additional Information

Dataset Curators

[More Information Needed]

Licensing Information

[More Information Needed]

Citation Information

[More Information Needed]

Contributions

[More Information Needed]

Downloads last month
82