Title: Doltics Radar
Author: doltics
Published: <strong>2026 年 6 月 3 日</strong>
Last modified: 2026 年 6 月 3 日

---

搜尋外掛

![](https://ps.w.org/doltics-radar/assets/banner-772x250.png?rev=3559124)

![](https://ps.w.org/doltics-radar/assets/icon-256x256.png?rev=3559124)

# Doltics Radar

 由 [doltics](https://profiles.wordpress.org/doltics/) 開發

[下載](https://downloads.wordpress.org/plugin/doltics-radar.1.0.0.zip)

 * [詳細資料](https://tw.wordpress.org/plugins/doltics-radar/#description)
 * [使用者評論](https://tw.wordpress.org/plugins/doltics-radar/#reviews)
 *  [安裝方式](https://tw.wordpress.org/plugins/doltics-radar/#installation)
 * [開發資訊](https://tw.wordpress.org/plugins/doltics-radar/#developers)

 [技術支援](https://wordpress.org/support/plugin/doltics-radar/)

## 外掛說明

Doltics Radar checks enabled WordPress flows before they complete. Admins can enable
protection for outbound email, submitted comments, and account registrations under
Settings > Doltics Radar.

Email protection uses WordPress core mail hooks when available. Doltics Radar also
includes a Mailgun compatibility adapter for Mailgun versions that replace `wp_mail()`
without calling `pre_wp_mail`.

### API Contract

The plugin sends a JSON `POST` request to the URL `https://radar.doltics.com/score`:

    ```
    {
      "text": "The sample text",
      "locale": "en"
    }
    ```

The API should return JSON using one of these response shapes:

    ```
    {
      "score": 0.95,
      "classification": "spam",
      "signals": {
        "token_count": 10,
        "unknown_word_ratio": 0.1,
        "spam_keyword_score": 0.61,
        "uppercase_ratio": 0.34,
        "exclamation_density": 0.3,
        "url_count": 1
      }
    }
    ```

Numeric scores are compared against the configured threshold.

## 安裝方式

 1. Upload this folder to `wp-content/plugins/doltics-radar`.
 2. Activate Doltics Radar in WordPress.
 3. Configure the plugin under Settings > Doltics Radar.

## 常見問題集

### What makes this spam scoring engine “smart” compared to traditional filters?

Traditional filters rely heavily on static keyword blocking, which spammers easily
bypass. Our engine uses a dynamic weighted scoring system. It extracts text, analyzes
standard language patterns, filters anomalies, and fact-checks content against localized
linguistic data. Instead of a simple “yes or no” flag, it combines multiple heuristics
to give you a highly accurate, nuanced probability score.

### How does the weighted scoring engine work?

The engine evaluates incoming text through multiple independent lenses (heuristics),
such as syntax anomalies, sender reputation, and localized language patterns.

### What does “localized language filtering” mean?

Spam looks different in every language and region. Slang, common idioms, and phrasing
vary wildly. Our engine doesn’t just translate text; it analyzes it against the 
specific rules and nuances of the local language context. This drastically reduces
false positives caused by regional dialects or informal business language.

### How does the engine “fact-check” text?

The engine cross-references extracted text and claims against verified localized
datasets to detect known phishing templates, fraudulent business pitches, and common
scam setups. If a piece of text closely mimics a known malicious campaign or contains
verifiable misinformation (like a fake lottery claim), its spam score increases 
accordingly.

### Does the text extraction process compromise user privacy?

Not at all. The text extraction and processing happen in real-time within a secure,
encrypted pipeline. We isolate the standard linguistic patterns and words required
for scoring, and we do not store the PII (Personally Identifiable Information) or
the contents of the processed text post-analysis.

### How does the engine handle sophisticated spam that uses intentional typos?

Because our engine extracts normal words and filters them against localized language
models, it excels at spotting “leetspeak” (e.g., replacing ‘E’ with ‘3’) or intentional
misspellings meant to trick basic filters. The system recognizes the underlying 
intent and penalizes the text for obfuscation tactics.

## 使用者評論

這個外掛目前沒有任何使用者評論。

## 參與者及開發者

以下人員參與了開源軟體〈Doltics Radar〉的開發相關工作。

參與者

 *   [ doltics ](https://profiles.wordpress.org/doltics/)

[將〈Doltics Radar〉外掛本地化為台灣繁體中文版](https://translate.wordpress.org/projects/wp-plugins/doltics-radar)

### 對開發相關資訊感興趣？

任何人均可[瀏覽程式碼](https://plugins.trac.wordpress.org/browser/doltics-radar/)、
查看 [SVN 存放庫](https://plugins.svn.wordpress.org/doltics-radar/)，或透過 [RSS](https://plugins.trac.wordpress.org/log/doltics-radar/?limit=100&mode=stop_on_copy&format=rss)
訂閱[開發記錄](https://plugins.trac.wordpress.org/log/doltics-radar/)。

## 變更記錄

#### 1.0.0

 * Initial release.

## 中繼資料

 *  版本 **1.0.0**
 *  最後更新 **4 個月前**
 *  啟用安裝數 **少於 10 次**
 *  WordPress 版本需求 ** 5.7 或更新版本 **
 *  已測試相容的 WordPress 版本 **7.0.6**
 *  PHP 版本需求 ** 7.4 或更新版本 **
 *  語言
 * [English (US)](https://wordpress.org/plugins/doltics-radar/)
 * 標籤:
 * [comments](https://tw.wordpress.org/plugins/tags/comments/)[email](https://tw.wordpress.org/plugins/tags/email/)
   [registration](https://tw.wordpress.org/plugins/tags/registration/)[security](https://tw.wordpress.org/plugins/tags/security/)
   [spam](https://tw.wordpress.org/plugins/tags/spam/)
 *  [進階檢視](https://tw.wordpress.org/plugins/doltics-radar/advanced/)

## 評分

這個項目尚無任何評論記錄。

[撰寫評分](https://wordpress.org/support/plugin/doltics-radar/reviews/#new-post)

[查看全部使用者評論](https://wordpress.org/support/plugin/doltics-radar/reviews/)

## 參與者

 *   [ doltics ](https://profiles.wordpress.org/doltics/)

## 技術支援

使用者可在技術支援論壇提出意見反應或使用問題。

 [檢視技術支援論壇](https://wordpress.org/support/plugin/doltics-radar/)