| --- |
| license: cc-by-nc-nd-4.0 |
| language: |
| - el |
|
|
| tags: |
| - greek |
| - wordpress |
| - blog |
| - articles |
| - news |
| - technology |
|
|
| size_categories: |
| - 1K<n<10K |
| --- |
| |
| # EELLAK and Other Articles |
|
|
| ## Dataset Description |
|
|
| This dataset contains articles collected from a multi-site WordPress network focused on technology, digital culture, public policy, education, and related topics. Each row corresponds to a single published article and includes the article body in Markdown format together with publication metadata. |
|
|
| The corpus spans multiple thematic websites and covers a long publication period, making it suitable for diachronic and domain-specific language analysis in Modern Greek. |
|
|
| The text is primarily in Modern Greek, with occasional English passages and bilingual content. |
|
|
| ## Dataset Structure |
|
|
| The dataset is distributed as a parquet file with one row per article. |
|
|
| ### Columns |
|
|
| | Column | Type | Description | |
| |---|---|---| |
| | `site` | string | Source website or subdomain | |
| | `id` | string | Article identifier unique within each site | |
| | `post_title` | string | Article title | |
| | `post_name` | string | URL slug | |
| | `post_date` | timestamp[us] | Original publication timestamp | |
| | `post_modified` | timestamp[us] | Last modification timestamp | |
| | `post_author` | string | Author identifier | |
| | `guid` | string | Canonical article URL | |
| | `content` | string | Markdown article body | |
| | `word_count` | int64 | Word count of `content` | |
| | `token_count` | int64 | Token count of `content` | |
|
|
| The pair `(site, id)` can be treated as the primary key. |
|
|
| ## Timestamps |
|
|
| Timestamps are stored as timezone-naive values representing local wall-clock time from the source WordPress installation |
