Semantic Resume

A simple HTML resume with semantic markups.

Features:

Single self-contained HTML file with HTML5 semantic tags;
Semantic markup using the vocabulary from semantic.org;
Generate from GitHub-flavored Markdown text with YAML header;
Basic theme support (try adding theme: modern to the YAML header);
Exportable to PDF;
Responsive layout.

Get started

Open https://semantic-resume.chunlianglyu.com/ and type in the editor. After you are done, click the Download button to download a self-contained single HTML file.

To export to PDF, use print and then choose "Save as PDF" as the Destination. Only tested in Google Chrome.

How it works

Markdown

A basic resume can be defined as in the following Markdown format:

---
website: https://chunlianglyu.com
github: cllu
---

# Chunliang Lyu

## Education

### [The Chinese University of Hong Kong] {2011 - 2016}

Research Area: Entity Retrieval, Natural Language Processing, Knowledge Graph.

[The Chinese University of Hong Kong]: https://www.cuhk.edu.hk/ (alumniOf)

If you have used Jekyll before, you would feel at home. It is just a standard Markdown text, with YAML header wrapped between two --- lines. Only the format of {2011 - 2016} may seems strange, which we will explain later.

HTML structure

For the above Markdown input, the generated HTML structure is like:

<main>
  <header>
    <h1>Chunliang Lyu</h1>
    <a href="https://chunlianglyu.com/">chunlianglyu.com</a>
    <a href="https://github.com/cllu">github.com/cllu</a>
  </header>
  <section class="education">
    <h2>Education</h2>
    <details>
      <summary>The Chinese University of Hong Kong, <time>2011 - 2016</time></summary>
      Research Area: Entity Retrieval, Natural Language Processing, Knowledge Graph.
    </details>
  </section>
</main>

As you see, we try to use several HTML5 semantic tags to hold content:

<main> to hold the whole resume;
<header> to holder, you know, the header;
<section> to hold a section with proper CSS class, like Education or Projects;
<h1> to hold the name, only in the first section; and <h2> to hold the title for other sections;
<details> and <summary> to hold an item in a section;
<time> to hold datetimes or time spans.

Semantic markup

Machines are still dumb and probably cannot understand the meaning of a piece of text. It may guess that the text inside h1 is probably a title, but is does not know the exact meaning of the text. Let't teach them, using the vocabulary from http://schema.org/Person. The marked-up version would look like:

<main itemscope itemtype="http://schema.org/Person">
  <header>
    <h1 itemprop="name">Chunliang Lyu</h1>
    <a href="https://chunlianglyu.com/" itemprop="url">chunlianglyu.com</a>
    <a href="https://github.com/cllu" itemprop="sameAs">github.com/cllu</a>
  </header>
  <section>
    <h2>Education</h2>
    <details>
      <summary>
        <span itemprop="alumniOf" itemscope itemtype="http://schema.org/EducationalOrganization">
          <link href="https://www.cuhk.edu.hk/" itemprop="url">
          <span itemprop="name">The Chinese University of Hong Kong</span>, 
        <time>2011 - 2016</time>
      </summary>
      Research Area: Entity Retrieval, Natural Language Processing, Knowledge Graph.
    </details>
  </section>
</main>

we mark the <main> element as http://schema.org/Person type, so a machine knows we are describing a person here;
we add itemprop="name" to the h1 element, so a machine knows this is the name for the person;
we add itemprop="url" to the website link, so a machine knows that url is our website;
we add itemprop="sameAs" to the GitHub link, so a machine knows the GitHub account there is the same person here;
the markup for the university is a bit of complicated, however it is not hard to understand. we tell a machine that we are an alumni of the defined EducationalOrganization, which has the given name and url link.

Why the bother marking up the HTML? And what the hell is schema.org? Well, believe or not, this kind of markup is already everywhere. Go to a YouTube page and check its source code. Search engines need to understand the Web better so they can improve the search result quality, that's why they join together to promote the semantic markup. From my point of view, after four years of digging information from text during my PhD study, I believe that Semantic Web is a better solution. As you can see, currently schema.org has limited support for describing a person, however, that is already a good start.

Develop

To start development:

Use npm install && bower install to install dependent node modules.
Run gulp --watch to start developing with a live reload server.

To deploy:

Run gulp --production to build production assets.
Run gulp deploy to deploy to GitHub Pages.

Main stack:

react for the UI; codemirror for the editor; marked for the Markdown parser;
scss for CSS preprocessor;
fontawesome for the lovely icons;
gulp for building tool;
webpack with babel so we can enjoy the awesome stuff from ES6;
BrowserSync for live reloading and multiple device testing;
mocha/chai for testing, and Travis CI so I can sleep tight in night;
GitHub Pages for hosting, and CloudFlare for CDN and free SSL.

There are plenty of resume generation projects over the Internet, free or commercial. These are some of them that I have surveyed. If you are not comfortable with hosting your resume, or dealing stuff like HTML and CSS, you would be better choose alternatives.

represent.io, paid service, nice interface, provides hosting and analytics
creddle
JobCV
ResumeRuby
TheResumator

Resume tips:

Credit

Initial versions of my resume are forked from JoyNeop/yart;
Sheets-of-Paper for the page style.

License

Beside the content of my resume, everything else is released under the MIT License.

greeve / Semantic-Resume