DIT@EKPA Undergraduate Curriculum

Structured, schema-validated representation of the undergraduate curriculum of the Department of Informatics and Telecommunications, National and Kapodistrian University of Athens (EKPA). The source is the official study guide (ODIGOS_SPOYDWN_DIT_EKPA_2025-26.pdf); this repo turns it into machine-readable data so it can be consumed by tools, drive UIs, and be reviewed change-by-change through Git.

Layout

schema/                     JSON Schema (Draft 2020-12), one per entity type
  course.schema.json
  program.schema.json
  specialization.schema.json
data/                       Source of truth — hand-editable JSON
  program.json              Academic year, institution, directions, graduation rules
  specializations/S1..S6.json
  courses/<CODE>.json       One file per course → clean, line-level diffs
scripts/
  validate.py               Schema validation + cross-file referential integrity
  build.py                  Bundle data/ → dist/curriculum.json
dist/
  curriculum.json           Generated single-file bundle for consumers
site/                       Static dashboard (Curriculum Atlas) — reads the bundle
  index.html  styles.css  app.js
.github/workflows/
  validate.yml              CI: validate on every push / PR
  deploy-site.yml           Build + deploy the dashboard to GitHub Pages

Website (Curriculum Atlas)

site/ is a dependency-free static dashboard that visualises the whole curriculum from dist/curriculum.json: headline figures (courses, ECTS on offer, offered-this-year, prerequisite links, instructors), charts (category mix, course load per semester, specialization depth, teaching-hours mix), an interactive prerequisite-flow graph (courses laid out by semester; solid links = required, dashed = recommended; hover to trace a chain), and a searchable/filterable course explorer with a detail drawer. Bilingual (EL/EN) toggle.

Deploy: the deploy-site.yml workflow validates the data, rebuilds the bundle, and publishes to GitHub Pages on every push to main. One-time setup: repo Settings → Pages → Build and deployment → Source: GitHub Actions.

Preview locally: serve the repo root and open the site —

python -m http.server 8000      # then visit http://localhost:8000/site/

(The page loads ./curriculum.json when deployed, falling back to ../dist/curriculum.json for local preview, so no copy step is needed.)

Usage

# one-time
python -m venv venv && ./venv/bin/pip install jsonschema

./venv/bin/python scripts/validate.py   # check schemas + integrity (exits non-zero on errors)
./venv/bin/python scripts/build.py       # regenerate dist/curriculum.json

Data model

• program — academic year, institution, department, degree title, the two directions (CS, CET), and graduation rules.
• specializations — S1–S6, each belonging to one direction (S1–S3 → CS, S4–S6 → CET).
• course — code, bilingual title, category, weekly hours (theory/tutorial/lab), ects, semester, direction, prerequisites (required vs recommended), specialization_roles (mandatory Υ / basic Β per specialization), offered_this_year, instructors.

Course code is the natural key; prerequisites and specialization roles reference it. validate.py enforces uniqueness and referential integrity beyond what JSON Schema alone can express.

Extraction

All courses were extracted from the PDF with scripts/extract.py (uses pdfplumber for positional table parsing). Re-run it any time the guide changes:

./venv/bin/pip install pdfplumber          # one-time, in addition to jsonschema
./venv/bin/python scripts/extract.py --dry-run   # parse & print a summary, write nothing
./venv/bin/python scripts/extract.py             # (re)write data/courses/*.json

It parses the course tables (pp. 76–83), the per-semester instructor schedule (winter pp. 88–89, spring pp. 91–93), and the free-elective lists (pp. 90 & 94), and adds thesis/internship from the prose. Specialization S1–S6 markers are read by x-position; codes are canonicalised to Greek; (ΔΠ) titles are flagged offered_this_year: false. Prerequisite type is read from character font/colour — bold-black → required (focused cycle), italic-blue → recommended (basic cycle), per the guide's own legend (p. 76).

Status

All 138 courses are extracted and validated (0 errors): 18 compulsory, 3 standalone labs, 11 direction electives, 4 projects, 3 general-education, 62 optional, 33 free electives, 2 thesis, 2 internship.

Prerequisite type is auto-detected from the PDF's bold-black / italic-blue formatting (48 required, 33 recommended).

Known caveats (worth a visual pass against the PDF)

• Instructors are filled where the schedule could be parsed (~86 courses); a few first-year/wrapped entries may be missing or partial.
• Codes for thesis/internship (ΠΕ1, ΠΕ2, ΠΑ1, ΠΑ2) are assigned by this dataset — the guide describes them in prose without codes.
• Full per-course descriptions (content, learning outcomes, assessment, bibliography) are not in this PDF; the guide hyperlinks them externally. Capture later via description_url if needed.