Network Science (2019/2020)

Evaluation (old version, before COVID-19)

Overview

Your grade is divided into 3 components:

There are no minimum grades in any of the evaluation components, but failure to deliver and present the project will result on a RFC evaluation (missing an evaluation component).

Mini-Tests and Mini-Assignments

During the semester you will be given two small individual assignments/homeworks (each worth 12.5% of your grade). You will have at least 2 weeks to solve some exercises and to apply some concepts in practice (potentially using a computer to analyze small datasets). You will send your answers to the professor's email.

You will also have two small tests (pen and paper) to be done during class (each worth 15% of your grade). Tests will be dimensioned to 1 hour and you will be given +30m of extra time.

These mini-tests and mini-assignments are individual. You will be given more detailed information closer to each evaluation date.
Tentative dates are the following:

  • 1st Mini-Assignment (12.5%) - available: 07/03 | due: 23/03
  • 1st Mini-Test (15%) - date: 24/03 | duration: 1h + 30m
  • 2st Mini-Assignment (12.5%) - available: 25/04 | due: 18/05
  • 2st Mini-Test (15%) - date: 19/05 | duration: 1h + 30m

  • Presentation/Reviewing an Article

    Type: individual presentation
    Date of presentations: 26th of May, 2020
    Time for each presentation: 10 to 15 minutes

    This is an individual assignment. Students should select a recent scientific article (year of publication >= 2017) about network science and should carefully read it so that they can present it orally to the other students and to the instructor. You should produce a small set of slides to assist you on the presentation.

    You should select a topic that interests you. The paper could be about a method, a model, an application, or any other subfield, as long as its core data being analyzed is of network type. When you have selected an article, please email Pedro Ribeiro so that I can validate your choice.

    You can look for articles on specific topics using academic search engines such as Google Scholar or Semantic Scholar.

    Here are a few examples of journals/conferences where you can find related articles:

    Since Network Science is a multidisciplinary area, there are many excellent related articles on conferences with a more broad range of topics. For example:

    You can also look at some "big" Network Science names and their recent papers:

    Project

    Type: individual or groups of two students
    Limit for project delivery (article): 2nd of June, 2020
    Limit for project presentation: 22nd of June, 2020

    Students will need to present the following deliverables:

    • Written article: 6 to 10 pages (using KDD double column format).
    • Presentation: 15 to 30 minutes, describing the work done

    The basic idea is for you to work on a small scale project in which you will analyze a network dataset using the concepts you learned in the course.

    You could either use already existing datasets, or create a new one (see the useful links section for some existing network datasets). In the same way, you can use already existing software tools or use your own code.

    You can also focus more on the analysis in itself (ex: showing insights gained), on the efficiency of the implemented algorithms (ex: showing execution times on different datasets), or on any other combination.

    You are strongly encouraged to speak with the professor so that you can validate your idea for the project before starting your work.

    Here are some example projects:

    • Papers Datasets: Take co-authorship network or a citations network and explore it. Possible tasks: differences between different areas and/or conferences, communities, importance of scientists, consider time, ...
    • Sports Datasets: Choose a sport and produce a results networks and explore it. Possible tasks: create ranking based on network, ...
    • Flights Datasets: Take all flights and explore airport importance, country importance, communities, ...
    • Wikipedia Datasets: Take a subset of Wikipedia and explore that subarea (ex: what is the most important article on a sub area, what communities are there, what is the difference on the network topology between different languages, ...)
    • Product Datasets: Take a co-purchasing or user-buying network and analyze it. Example tasks: predict links, communities, ...
    • Review Datasets: Take an user-product review dataset (ex: movies) and analyze it. Example tasks: predict links/scores, communities of users and/or products, ...
    • Biological Datasets: Take a biology network (ex: protein interaction, gene regulation, brain networks, ...) and explore it. Example tasks: discover characteristic patterns, ...xs