About Presentation

Presentation for academic purpose


  1. Fluent transition between subsections. Remember to note the bigger picture from time to time.

  2. For the literature review part, remember to discuss the pros and cons of existing work with an objective view.

  3. For presentation, don’t fill it with your whole paper. Presenting 30% of the content clearly is better than pouring 100% of blurry concepts.

  4. For PQE/survey, spend a bit more time on the challenges that are faced in your topic area.

Presenting visualization

  • What’s the story you find from the data using your visualization tool?

  • Rehearsal: print slides so that others can mark and give feedback

  • Last slide: take home message?

Arrangement of the pre

  • Time arrangement!

  • Concrete case study, story!

  • Redundance incase some may lost

  • Motivation?

  • Related work (a few names and screenshots to show respect)

Present to outside researchers

  • What is the problem? Why trivial solution doesn’t work?

    • analytics tasks
  • Why is it important?

  • Why is it challenging? E.g.

    • Spatio is easy, temporal is easy, but spatio-temporal is very hard…
    • Scalability, when data become to large scaled, very high dimension
  • How existing work improve the state-of-the-art? Their contribution?

    • Form relations between works
  • Unsolved issues?

Present to normal people

  • How to impress them? Know your audience!

    • if the audience are investors, show them what you have done with numbers and cases.

Three stories

  1. What is the domain problem, how visual design solve it. Visual channels.
  2. Scalability, how people solve it.

How to present visualization?

  1. Throw questions. What the visualization is trying to solve?
  2. What

Industry visualization

Smart City (Face to governments)

4 needs

  • 数据获取
  • 数据的融合 data fusion. Eg., Security monitoring 态势. Traffic, and etc. 5要素组织?大数据的展现,移动终端的展现。 Multi-level.
  • 数据展示
  • 数据服务


  • Recomend the result/method directly. Reduce user’s operation. KEY Problem: How to define the scene? Need a platform to support the working mechanisms.

Key issue:

  • API设计, 粒度。 Their solution: 1. 模版
  • 政府数据在私有云,只能在内部学。数据的noise