Amy Guy

Raw Blog

Wednesday, November 28, 2012

Advice: willful misunderstanding, and audience as individuals

Picked up on some great advice during an afternoon-long course entitled 'How to do an Informatics PhD' a couple of weeks ago.

All through my undergraduate, and probably High School as well, I was told that when writing assignments I should treat the person marking it as if they don't know anything at all about the subject.  They're stupid.  Leave nothing unexplained.  Of course, we were often also told to 'keep things concise' and usually had to do this under the constraints of page or word limits.

Problematically, some people could have interpreted this as an opportunity to try to pull the wool over their marker's eyes, or baffle them with science.  Typically, the marker will know something about the subject, so that's probably not a great tactic.

I see where this advice comes from of course, and a couple of weeks ago I heard it phrased in a different way, that makes far more sense.

The examiner will be knowledgeable about the subject, but given to willful misunderstanding of what you're trying to say.

I think that gives a much better guide to how and when to explain things.  And also makes them more of an enemy to be conquered, than an inconvenient fool to be deceived.

And whilst I'm recounting advice I like, here's something that has stuck with me since 2010, from my manager at Google at the time.

When presenting to a large audience of people, don't think of them as a crowd.  Think of them as many individuals.  Make your presentation as you would to a single person.  It just so happens that there are lots of single people there all at the same time.

Since letting that sit in my subconscious, nerves before giving a presentation have shrunk to negligible levels.  It may be that over the past two years I've become more confident anyway, but I used to be thoroughly terrified of standing up in front of even a classroom of people the same age as me.  I wouldn't talk out in lectures for the most part of my undergraduate, and harbored a gut-wrenching fear of being picked on to answer a question.  As did most people, I imagine.

I make sure to actively observe how I feel when watching someone else present.  I look at other people in the audience too, and note the attitude and techniques of the presenter.  (Consequently I probably leave having no idea what the presentation was about).  The results are usually that a majority of people aren't listening properly.  A vast proportion certainly aren't angrily judging the presenter's every twitch.  Things people notice and get upset about include:

  • If they can't hear you.  
  • If you're just reading off slides, particularly if you try to act like you're not.
That sure is a short checklist of things to avoid.  There must be more that can go wrong.  Reasons people will stop listening, include:
  • If they can't hear you.
  • If you're just reading off slides, particularly if you try to act like you're not.
  • If they're not interested in what you're talking about. (Pro tip: make them interested).
  • If you're not making any sense at all.
  • If your slides are more interesting than what you're saying. (I really like presentations without slides.  So long as the speaker is engaging, of course.  Slides with lots of words are a definite negative, in my book though).
And some of my personal nitpicks include:
  • Drawing attention to a mistake by apologising for it.  From being on both sides of this situation, it usually feels right to do so at the time, but until you do, four fifths of the room won't have noticed, and the fifth that did will forget within the next few seconds.  Point it out, and everyone will remember.
    • If you say something wrong, just correct yourself and move on (but don't leave it uncorrected, this is usually noticeable).
    • If it's a technical problem, keep talking whilst it gets sorted.  This boils down to not relying on technology to keep your presentation interesting.  I am aware that there are some situations where this is impossible.
  • Weak intros and outros.  I've been guilty of both of these.  I plan to pay more attention to upcoming talks in order to fix this.  It's something I always forget to notice.  But for now:
    • Starting with filler words, like 'So, ...', 'Right then...' or 'Okay, ...'.
    • If you've been introduced, you don't need to repeat it, especially not in a way that draws attention to the fact you're repeating it. 
    • Make it clear when your talk is done.  Don't trail off with '...and that's that then.'  'Any questions?' is usually okay, but only if it follows a distinctive final sentence.  Jumping to that from what feels like half way through a paragraph is a bit rubbish in my head, but in all honesty will probably go unnoticed by the audience.  Except me.

Fortunately, my days of feeling agonisingly self-conscious whilst presenting are long gone.  On top of that, I find doing as little 'rehearsal' as possible boosts the natural fluidity of a presentation, and in turn my confidence.  If I haven't rehearsed, there's nothing to forget to say (and suddenly forgetting what comes next is the biggest killer of flow, something I discovered during French oral exams).  That only works if you're very familiar with the subject matter.  And if you're not, you probably shouldn't be presenting about it.

Disclaimer: I'm very early in my academic career, and haven't presented a whole lot.  The biggest audience I've talked in front of was about 120.  Despite the theory, I suffer from having neither a naturally loud voice, nor a naturally beaming expression.  So all round, I'm probably not very good.

Monday, November 26, 2012

Notes about Semantic Web tools for online communities

K. Faith Lawrence & Dr. Monica Schrafel (2007)  Amateur Fiction Online - The Web of Community Trust: A Case Study in Community Focused Design for the Semantic Web. Intelligence, Agents, Multimedia (IAM) Group, School of Electronics and Computer Science, University of Southampton.

NB. Need to read her full thesis, of the same name.  Will probably clear up some of the questions I scribbled whilst reading the paper.

Finding out if Semantic Web tools can be brought to hobbyist groups on the Web.
  • Uses the online fiction community; suggests they could benefit from:
    • improved searching
    • improved meta data
    • automatic recommendations
    • trust webs
    • personalisation.
  • A HCI project, so usability tests and comparisons with current systems are key.
Related work
  • Community centered design
    • to determine user needs - through continual interactions and user studies.
    • to consider how reader-facing apps present themselves and particular community
      • responsibility of being a portal - need clear affordences and points of failure.
  • Trust and Semantic Communities
    • The semantic web “provides a common framework that allows data to be shared and reused across application, enterprise, and community boundaries.” - Tim Berners-Lee, James Hendler, and Ora Lassila. The semantic web. Scientific American, May 2001.
    • Do we trust:
      • metadata
      • data
      • mechanism by which data is returned
      • person requesting data?
    • Many definitions of trust.
    • Jennifer Golbeck's trust onotology to go with FOAF (Jennifer Golbeck, Bijan Parsia, and James Hendler. Trust networks on the semantic web. In Proceedings of Cooperative Intelligent Agents 2003, 2003.): ratings of 1 - 9 for trust of associates.  Extended for FicNet (this paper) 
    • Here, trust: "the expectations that arise that an individual will not act in a way that is detrimental to another individual or community."
      • I might need to expand that for my stuff... maybe... maybe this will do.
Case study
  • Community predates the Internet. (duh)
  • Necessary to get opinions from people outside of the amateur writing community, because it is broad. Such as parents/guardians of members.
  • Questionnaire
    • General information
    • Reading habits
    • Community involvement
    • Access and distribution of materials
  • Questionnaire distributed by:
    • requests to archives to pass on to members
    • LiveJournal
    • emails to specific interested parties
    • mailing lists / bulletin boards of relevant special interest groups.
  • In two weeks, 1116 responses, from 30 countries.
  • Used to inform ontology design for FicNet, and OntoMedia.
Fan Online Persona (FOP)
  • Extension of FOAF, tailored for needs of online readers and writers.
  • foaf:person -> fop:persona
  • Separates environments for on and offline
  • fop:NomDe - context for name
  • Illusion of anonymity is fundamental to fanfic community (who are a large part of online amateur writers)
  • Most authors have one or more pseudonyms.
  • 80% said email address is the most personal information they should be asked for.  ('of the 80%, 15% said no personal information should be requested from anyone - does this make sense? Do they mean other than email address?  but that's the 80... If they're in the 80, they can't be part of that 15...)
  • Privacy is the main thing holding back FOAF (Joseph Smarr. Technical and privacy challenges for integrating foaf into existing applications. Presented at 1st Workshop on Friend of a Friend, Social Networking and the Semantic Web, September 2004.)
  • Personas aren't meaningless, because people become very attached to them, and only create new ones for specific reasons (says who? No citation..)
  • Expands foaf:document and foaf:groups
  • Creation, exchange and review of works is the point of these communities.
  • FOP dismisses FOAF info like work and school as irrelevant or potentially dangerous.
    • [Me]  I think the on/offline divide won't be so extreme for many amateur film makers (another story for consumers) because often their faces are in their movies... Also anecdotal evidence from my own experiences that I'm open to having proven to be a minority.  Actors vs characters is an interesting distinction too.  One amateur film maker can have many personas, even across one channel of output.
  • Options for FOP determined through long term study of metadata commonly attached to works. (Something I can do, too).
  • FilmTrust by Golbeck (just joined, it was closed last time I looked).
    • Could be prettier... but 2169 members!
    • Visualisations of the network - I need to get good at this.
  • Reader has to trust info from author falls within a certain level of accuracy.  In amateur writing, it is more acceptable to be over cautious than lenient.  Differing standards of acceptable content.
    • Less trust is lost if a story is underrated than overrated (resulting in disappointment)
    • A minority mislabeling work has a big effect on reputation of an archive/community (HelpingHands community members. A place to pitch in and help - a website creation resource and project. LiveJournal Community, 2005.)
  • Writer has to trust reader to make the right decision.
  • FicNet has a more specialised trust system than Golbeck's.
    • Largest contention in this field is adult material and younger readers (debated because this contrasts with IRL - no restricted areas in book stores, or suitability rating scheme for books).
    • Initially focussed on age.
    • Personas could vouch for each other.  Creating fake personae to validate another wasn't worth payoff?  Non malicious statements of distrust?
    • How to integrate trust and distrust webs?
  • Ontologies developed, ready to be used by applications!
    • Ontologies will be continually refined.
    • Now designing applications.
      • Using info already gathered via quesitonnaire, re: UI, functionality.
  • Integrate with OntoMedia to describe works, and link works with people.

=> How did they get to talk to the parents of younger users?  Did they ask the members to put them in touch?  That doesn't seem like a realistic expectation to have, to me..

Week in review: Pancakes & Project management

19th November - 25th November

I read two papers about ontology development methodologies.

I read two articles by Bennett Haselton about decentralized social networking, which happened to pretty much sum up and beautifully articulate everything about that that has been floating in my subconscious for a couple of weeks.  I saw links to them in the latest Circumventor email, which I've been subscribed to since High School for bypassing the internal blacklist, and remain subscribed to because the jokes at the end are always laugh-out-loud funny.

I attended an all day course entitled 'practical project management for research students'.

  • It was attended by a diverse bunch of seemingly really lovely people.
  • The two ladies running it, from the IT Project Management department in the University, were lovely too.
  • The stuff covered was all obvious, common sense stuff (and pleasantly the organisers didn't try to claim otherwise) - but sometimes it's helpful to have it all written down and waved in your face.  And structured, in particular.  Made me actually focus on thinking about organising my project.  The main thing I hadn't much considered, even subconsciously, was formally identifying stakeholders for a project and their relative interest/power in the project.
  • There are a bunch of tools at to aid in project management.
  • It prompted me to do these week-in-review posts, as I realised I haven't been recording properly everything I've been doing (an overview of my time goes on my calendar, but no detail).
  • The sandwiches weren't great, but fortunately when I got back to the Forum there were massive slabs of chocolate cake left over from some event.  I love the Forum.
I booked a place at the 1st International Open Data Dialogue in Berlin, and necessary flights.
  • Despite the short notice, it worked out logistically because I need to be in London on the 7th anyway, so I can simply go to London on the 4th, fly to Berlin from there for the 5th and 6th, and back to London for the 7th.
  • I'm particularly looking forward to "Open Statecraft: Openness as a Means (not an End)" by Philipp Müller, "The Open Data Movement vs. Business Models - is this a Contradiction?" by Dr. Peter A. Hecker, "Linked Open Data @ W3C-Vocabularies, Working Groups, Usage Scenarios" by Prof. Felix Sasaki, "The potential of Open Data for improving urban sustainability" by Dr. Marianne Linde and "Towards Trustworthiness: Establishing Transparency with Open Information Flows" by Dr. Edzard Höfig.
  • I'm also looking forward being in Berlin again, even if it is just for one evening, and I'll probably be too exhausted to appreciate it.
Ontologies with a View took place at a different place and time to usual.
I started preparing for presenting at Digital Methods as a Mainstream Methodology in London in a couple of weeks.
  • I scribbled lots of notes.
  • I skimmed a few papers by organisers/speakers but didn't read any in detail yet.  Mostly stuff about analysing data gathered from comments, tweets etc. 
  • There will be more about both of those things next week, I imagine.
I made a plan for the two weeks following the 20th.
  • It mostly consists of finishing my Digital Methods preparation.  I have a lot of non-PhD related things to do as well, plus lots of travelling.  Also graduation from my MSc, and subsequent parental visitation will get in the way.

Wednesday, November 21, 2012

Notes about ontology creation methodologies (2 papers)

Yesterday I unexpectedly read two whole papers about ontology development methodologies.  They were open in tabs I don't remember opening, but presumably did so during our weekly Ontologies With A View meeting last Friday.  There are still a bunch more tabs open with papers or articles about the same thing, so maybe I'll read those later..

The notes are here more or less as I scribbled them down whilst reading, and I haven't expanded with any analysis or discussion as of yet.

Notes in purple are things I intend/need to investigate further; colour-coding is just for me, really.

Jean Vincent Fonou-Dombeu & Magda Husiman (2011)  Combining Ontology Development Methodologies and Semantic Web Platforms for E-government Domain Ontology Development.  International Journal of Web & Semantic Technology (IJWesT) Vol.2, No.2, April 2011 

  • Start by describing ontology in a human-readable way, then turn to RDF (etc) to be machine readable.
  • Says there's not sufficient practical research around existing technologies or ontology development guidelines that would allow non-experts in e-government domain to make ontologies.
  • Uses framework from Uschold & King (see later in this post) to describe ontology - technique used here should be platform independent.
  • Then uses UML to semi-formally represent ontology
  • Uses Protege and Jena to convert to OWL and RDF
  • Paper's goal is to produce guidelines for e-government developers to create semantic content AND strengthen adoption of Semantic Web technologies in governments (particularly developing countries).
  • Outlines RDF, OWL, Protege and Jena (described as leading platforms; mentions other platforms: WebODE, OntoEdit, KAON1, Sesame).
  • Very critical of other literature; either ontologies have been produced but no practical information given; they've been developed with proprietary platforms; or they're only conceptual and don't say how they could actually be constructed with existing technologies.  other studies have not focused on a methodological approach, which means nothing is easily repeatable.
  • Detailed comparative studies of methodologies in:
    • M. Fernandez-Lopez, “Overview of Methodologies for Building Ontologies, ” In Proceedings of the IJCAI-99 workshop on Ontologies and Problem-Solving Methods (KRR5), Stockholm, Sweden, 2 August, 1999. 
    • H. Beck and H.S Pinto, “Overview of Approach, Methodologies, Standards, and Tools for Ontologies,” Agricultural Ontology Service (UNFAO), 2003. 
    • C. Calero, F. Ruiz and M. Piattini, “Ontologies for Software Engineering and Software Technology, ” Calero.Ruiz.Piattini (Eds.), Springer-Verlag Berlin Heidelberg, 2006.
  • Case study: Ontology for monitoring development projects in developing countries (OntoDPM)
    1. Create with Protege:
      • class heirarchies
      • slots
      • domain and range of slots
      • Based on the UML
      • Saved as OWL
    2. Then put content in RDF with Jena

Mike Uschold and Martin King (1995*) Towards a Methodology for Building Ontologies. Workshop on Basic Ontological Issues in Knowledge Sharing, IJCAI-95.
  • Steps:
    1. identify purpose
    2. build ontology
      • capture
      • coding
      • integrating existing ontologies
    3. evaluation
    4. documentation
  • Purpose
    • Many ontologies are intended for reuse
    • Should survey purposes to clarify options for future projects
  • Building
    • Capture
      • Identify key concepts and relationships in domain of interest
      • Produce unambiguous text definitions for these
      • Identify terms for these
      • Agree on all of the above
    • Coding
      • Explicit representation of conceptualisation in a formal language (choose a language)
    • When can capture and coding stages be merged?
    • Differences between building ontology and creating a general knowledge base (thinking about methodology will help with this)
    • Integrating (during either or both of above)
      • Work must be done in agreement between communities
      • Make explicit all assumptions underlying an ontology
  • Evaluation
    • Judge against requirements specification (and/or)
    • Judge against competency questions (and/or)
    • Judge against real life
    • This paper looks at knowledge base systems, and adapts for ontologies.
  • Documentation
    • Desirable to have established guidelines for documenting
    • Main barrier to effective knowledge sharing is inadequate documentation
    • ALL important assumptions should be documented
  • Case Study
    • Main emphasis is on capture phase
    • Initially:
      • define ontology (Gruber)
      • identify users and usage (initially abstract, then clarify with real life)
      • choose language (Ontolingua was chosen)
      • choose method for capture - BDSM (IBM) supported by others:
        • KADS
        • IDEF5
        • OO Analysis and Design techinques
        • Gruber's principles for ontology design
    • Categorisation is fundamental to the human condition (Lakoff)
      • Not heirarchical, but:
        • GENERAL
             BASIC  ->  primary with respect to knowledge organisation
        • eg.
                SUPER:   Animal    /   Furniture
                BASIC:    Dog        /   Chair
                SUB:        Retriever /   Rocker
        • Certain concepts used subconsciously, rather than understood intellectually.
          • These have a more important psychological status.
        • Therefore paper uses middle-out approach to capture terms
          • (bottom-up = too much detail unnecessarily,
          • top-down = risks imprecision)
        • BASIC concepts first because:
          • most important
          • used to define non-BASIC terms
          • increase clarity, especially for non-technical use
          • backed by BSDM experience of paper author
  • Scoping
    • Brainstorming
    • Consult corpora if there aren't enough domain experts to brainstorm
    • Grouping
      • structure terms into naturally arising sub-groups
      • collate synonyms
      • consider things that might refer to each other
  • Meta-ontology
    • Don't commit too early, can restrict thinking.  Let concepts and relationships themselves determine requirements.
    • Be consistent.
    • Use technologically neutral language ('thing' vs 'entity').
    • Start with areas where there's most overlap.
    • Work from basic terms to more abstract ones within an area.
  • Producing definitions
    • Agreeing on definitions (varying degrees of problems)
    • Handling ambiguous terms
      • clarify ideas without technical terms
      • use a dictionary!
      • label definitions, eg. x1, x2
      • determine most important concept
      • choose a term, avoiding original ambigious one
    • Avoid new terms
    • Terms get in the way (peoples' preconceived ideas) - concentrate on underlying meaning and concepts.

Friday, November 09, 2012

National Novel Procrastinating Month

It's day nine, and I'm on two thousand, eight hundred and seventy words.

A quick calculation might tell you that that means I'm quite behind schedule.  This may be my worst year yet.  There's still plenty of time to get back on track though!  Right..?!

I've only spent any time writing on about three or four of those nine days so far.  But I have been to a conference, organised some SocieTea events, read bits and pieces related to my PhD, cleaned my flat, watched a few episodes of Arrested Development, learnt some new crochet stitches and started crocheting a hat, and baked a lot.

I did meet the Edinburgh NanoBeans and had a great time at the write-in in Pulp Fiction last Wednesday.  We may have spent more time collaboratively developing the backstory of Pedro the Guide Bear (a troubled young grizzly attired in an Elvis costume and boater hat who constantly struggles against his estranged father, Yogi, the leader of an organised crime syndicate) than actually writing our novels though.

I have learnt one particularly important thing this year, that's never come up before.

Talking ideas through with other people is really useful!  

Last Sunday, Beth helped me explain the absence of a main character's mother and fix a potential looming plot hole with one fell swoop.  Telling Kit about the various civilisations and layout of the land in my world allowed him to pick holes and question things, raising, and partially solving, some things that didn't make sense or yet more potential looming plotholes.  And Caitlin (a new NanoBeans writing buddy) pointed out that just because a character had been anticipating reading a letter for the last thousand words, didn't necessarily mean the letter had to contain anything interesting... it could be a disappointment to the character... which helped, as I hadn't figured out what the letter said, and all of a sudden the character was opening it.

I sure wish blogging about Nano counted towards the word count.

Monday, November 05, 2012

Remediating the Social #elmcip

I spent the last few days in Edinburgh College of Art, helping out at the Remediating the Social conference.  I was in charge of making sure everyone's microphones were on, and slides were being projected, which turned out to be more work than anyone anticipated.  Only minor hiccups occurred though, usually when I unplugged something I shouldn't have by accident.  I couldn't have done it without my glamourous assistant José, who was the master of fiddling with Macbook screen resolutions to make them play nice with the projector.

More importantly, I saw some super interesting talks, and met and talked to some fantastic smart people about electronic literature, and other things.

I also presented about Palimpsest, in front of the biggest audience I have ever talked in front of.  Go me.

Videos of everything from the conference are here.

On the last day I implemented an idea that had been kicking around the back of my mind for a while, which was the Uninformative Twitter Wall, or Twitter Squares.  It's nothing particularly complex; it uses jQuery and probably has memory leaks.  I'd love for people to help themselves to the code and improve it. Converting a hash of a tweet text into a hex code, I generated coloured squares for the results of a search term.  If the feed you choose is updating a lot, then the squares move around quickly and it looks pretty funky.  If there are only occasional new tweets, then it looks less exciting, but is still equally useless for seeing what people are saying.  (Unless you hover over the squares).  That's okay though, because it's Art.