Metrics to Consider

  1. Similarity of nodes, specifically structural equivalence and cosine similarity
    • Cosine Similarity: measure of structural equivalence on undirected networks
      • Need a way to normalize count of common neighbors across graph: Pearson coefficient (compare to expected value in random network)
      • Pearson Correlation Coefficient : where is the indicator random variable that a randomly chosen node is the neighbor of node .
    • Direct Application: Measure of similarity that takes into account structural equivalence as well as similar properties/characteristics, that all evolves over time
      • What marriages are similar to X?
      • What marriages show a similar trend over time to X?

To Do

  • Nauvoo Data
    • Add timestamp to the database tables (for when last edited, default to NOW())
    • Have the save button directly write back to the db (done)
  • Networks
    • Look at the evolving network and how to conceptualize that for the social groups (AQ, etc)
      • is that a chord diagram of groups?
      • Networks of those things with connecting people, or something else?
    • Measure of connection between marital graph and social graph
      • how does it change if the marriage graph is matriarchal vs patriarchal?
      • Measure of the interaction between the graphs
    • Find other applications and domains for this marriage vs organization kind of comparison
      • SNAC
        • Chord is a document with people coming into it. People who are there because of context or because of content of the record. So, authors and consumers vs referenced?
        • Or, chord is a box (of documents, a collection) and people are connected to that box (by being related to that collection)
        • Vs (this would be the organization) Institutions where the people and documents are connected to their respective institutions.
        • That is, (question) how documents interrelate to institutions (like marriages and orgs)
  • Background
    • Let Worthy know about:
      • Course material from Bruno
      • Results of reading papers, such as Characterizing Continuous Time Random Walks
  • What is the Time Varying Network for organizations?
  • Make notes on boundary conditions:
    • For marriage TVNs, we add boundary nodes (for parent marriages or child marriages)
      • Do we use one placeholder node?
      • Do we leave them out of the metrics we create?
      • Do we have one per boundary edge?
    • For snac, if someone is mentioned in only one document, we don’t have 2 ends of that edge in the network
    • What are our options?
      • Null nodes
      • single null node
      • dangling edge
      • self loop
      • edges that go to alternating bipartite graphs
    • What do the measures say in each case?
      • What people are connected to which other people?

To Discuss

  1. Is there any update on the requirements from SIS? Do I need another class or math course?
  2. Presidential Fellowship in Data Science with Adam. Should I do it?
  3. Data Entry page is fairly complete. Few last straws to fix and a workflow to attach to the “Save” button
    • Do we want to use an open authentication method? Or write our own?
    • Do we want changes to be written to another database pending approval? Or directly modify our database?
  4. Have MSSQL data dump. Has Worthy done conversion before?
  5. Added the working set lists to the DB, but haven’t added those individuals to it yet.

Notes

  • Nauvoo Data
    • Shayne will pull the SQL out of the MSSQL *

Understanding the science of networking

  1. What networking is
    • relationship building
    • genuine interactions with other people who share some similarity with you
    • exchange of information, ideas, support
  2. What networking is NOT
    • Fake, insincere, useless
    • a bother or intrusion
    • a skill that requires a particular personality type to be successful
    • only a way to get a job

Approach to networking

  • Motivations for networking
    • What do you want to get out of this, a job, a connection?
  • Managing expectations
    • Networking looks different for each person
    • Hardest part is getting out there to begin with
  • Timelines/results
    • How do you kno if you are successful with your networking?
  • Common scenarios for networking
    • Networking reception/event
    • Information sessions
      • Once you get a business card, send a follow-up thank you email with details of what you talked about with that recruiter/rep, and tell them you want to know more about their work.
    • Informational interviewing

Mastering the Art: Elevator Pitch

  • Your story
  • Think about who you are talking to/who the audience is when deciding what to include
    • What is relevant as it relates to them?
  • Who are YOU? How did you get to this point? Where do you hope to go?

Mastering the Art: Verbal Communication

  • Questions and conversation starters
    • What are some great ways to start a conversation?
    • What questions should I stay away from?
    • How do I jump into a conversation?
    • How to be assertive and confident
    • Do I need to know about sports entertainment, news, etc?
    • Appropriate/inappropriate language
      • Be a little more professional than the person you are talking to

Mastering the Art: Nonverbal Communication

  • Nonverbal communicatino/etiguette
    • Attire: what are your clothing choices saying about you?
    • Approach everyone with confidence and smile!
    • Use an open stance posture
    • Eye contact
    • Graciously invite others to join your conversation
    • Don’t linger too long
    • If there is food served, remember that it should not be the focus
    • Leave your right hand free for handshaking

Thank you and follow up

  • ALWAYS send a thank you
  • You are not really networking if you never follow up after the first (or second) conversation

Using the science to master the art

  • Quality over quantity
  • Information seeker (not a job seeker)
  • Never miss an opportunity to say thank you
  • Listen
  • Be genuine and true to your word
  • Impressions matter
  • Mentors
  • Follow-up

Publishing Venues

Venues with immediate deadlines

  • Family History Technology Workshop at BYU
    • Mentioned by Luther
    • February 10, 2015
    • Abstract (1-2 pages) due by January 19, 2015 to fhtw@internet.byu.edu
    • Call for Papers
  • SDM Doctoral Forum 2015
    • Deadline for Abstract February 2, 2015

General Venues for the future

  • Evolving Networks
    • SIAM International Conference on Data Mining
      • http://www.siam.org/meetings/sdm11/
    • Social Networks, Journal of Elsevier
      • http://www.elsevier.com/locate/socnet
    • ACM KDD Conference on Knowledge Discovery and Data Mining
      • http://www.kdd.org/kdd2015/ (Aug 2015)
    • ACM Social Network Systems
    • ACM/IEEE Advances in Social Network Analysis and Mining
      • http://www.asonam2014.org
      • Aggarwal was keynote speaker in 2014
    • Ad Hoc-Now (2012 Time-Varying Graphs paper was here)
      • http://www.adhocnow.net
    • Modeling and Mining Temporal Interactions
      • http://m2ti.weebly.com/
  • Visualization
    • IEEE Visualization
    • Social Networks, Journal of Elsevier
      • http://www.elsevier.com/locate/socnet
    • Journal of Social Structure
      • http://www.cmu.edu/joss/
    • American Journal of Sociology
      • http://www.jstor.org/action/showPublication?journalCode=amerjsoci

Notes by David Proffitt

  1. Preparing
    • What do you want to say?
      • Big issue being addressed
      • Specific question under discussion
        • What you did
        • What you found
        • What it means
        • Narrow topic, not all about you!
    • Who are you talking to?
      • For cog lunch, it would be a non-specialist
      • Don’t assume your professors are specialists, it’s okay to talk to them
    • Make slides
      • Good order for slides
        • Title
        • Issue
        • Background
        • Research question
          • What’s missing from the background. This is your research question.
          • Give the answer up front: the research I’ve done is going to show that $X$. That way the audience can determine whether it actually supports it throughout the presentation.
            • This is highly debated, even in this talk.
        • Design and Methods
        • Results
        • Conclusions and Implications
        • Questions
      • He suggests only using black text on white background. White on black is really white on gray when projected.
        • “Will annoy any visual scientist in any talk” if you use use anything other than black on white.
        • Don’t decorate your slides (logos of your lab, etc)
        • Purpose of the slide is to be legible
        • Only use color to highlight certain aspects
      • Font sizes
        • Don’t use anything smaller than 24pt.
        • The eyes don’t age very well
      • Avoid too uch bulleted text
        • Reading is obligatory (as an adult, you can’t avoid reading). See the Stroop Test.
        • Don’t put your lecture notes in your slides!
        • Don’t put text in your slides that you’re not going to say aloud. It produces cognitive fatigue (audience is struggling to decide between reading your slide and listening to what you’re saying).
        • What to do with bulletted text
          • Great for creating initial drafts, but then replace it with pictures, charts, etc.
          • Flowcharts and tables are great replacements for bullets
      • Data slides
        • Label Axes
        • Include Error bars
        • Then, when you are talking about it, describe what everything means. (axes)
      • Tie your future directions to your talk
      • The last slide should be a question motivator. You want people to ask questions, so don’t show your collaborators.
        • Put something up that helps them think of questions
        • If no one thinks of questions, try to come up with a softball questions to get other people going asking questions
    • Practice
      • Present to a friendly audience (familiarity, timing)
      • Present to a group in another lab (that don’t know what you’ve alread done)
      • Take your talk for a walk: find a park or somewhere quiet, and walk for hours before the talk and rehearse while walking.
        • Know the talk cold
      • Don’t write your whole talk and read it. It might help to write the first paragraph just in case.
        • People can tell if you’re reading
        • After the first paragraph, you’ll be into it
  2. Presentation
    • Know your audience
      • They might be preoccupied, inattentive, unfamiliar with your project, easily bored, easily amused.
      • At any given time, 1/3 of your audience isn’t listening
        • For important concepts, say it more than once
    • Face your audience, not your slides
      • Don’t use a laser pointer if you’re prone to nervousness (the dot will be moving as you shake, which is distracting)
      • Look into eyes, not behind them
    • Watch your audience for signs of life. Are they getting too bored or getting lost?
      • Do this when you teach. If you’re losing them, go back and ask if you need to go over it again.
    • Be yourself!
    • Don’t use cartoons, but if you do, walk the audience through it and laugh at it yourself
    • Questions
      • When someone asks a question, listen to the question
      • Understand (rephrase and repeat the question back, or ask the person to repeat it)
      • Answer the question
      • Not knowing the answer is okay. The ones that you don’t know are probably not directly related to your talk, but only tangentially related.
        • Take this as an opportunity to throw the question back at the person, “how do you think that would play out?” Let them develop the implications
  3. Evaluation
    • Get feedback on your talk
    • Pay attention to what works in other people’s talks as well as your own
    • Go to every job talk that you can, and don’t make the same mistakes
    • Steal ideas whenever possible
      • Analyze and emulate other people who give talks that you admire, but BE YOURSELF!
      • Find people who you feel empathy with, and steal those