Author Archives: navatintarev

Research visit to California

goletaI (Nava) have just come back from a fantastic trip to California, supported and enabled by the generous SICSA PECE grant to visit the Four Eyes Lab at the University of California Santa Barbara (UCSB) and Dr. John O’Donovan.

During my time at the 4Eyes lab I learned more about interactive visualizations, and we designed a user interface that contributes to the discovery of novel and relevant content, and improves perceived transparency and control. With Jay Byungkyu Kang, Tobias Höllerer and John O’Donovan at UCSB we worked on a submission for the workshop on interfaces and human decision making for recommender systems (IntRS) in conjunction with the ACM Recsys conference in Vienna. We are currently improving on this work in a system that uses live twitter data.

system-mockup1

The submission, titled Inspection Mechanisms for Community-based Content Discovery in Microblogs”, was accepted for publication.

I also visited and gave invited talks at Yahoo! Research (Sunyvale), and InTouch Health (Goleta).

Leave a comment

Filed under Uncategorized

Invited talk at Centre for Human Computer Interaction Design London

On the 13th of February Nava gave a talk at HCID at City University London: Evaluating the Visualisation of Plans in the SAsSY Project (Talk slides)

Screen shot of first slideAbstract: The Scrutable Autonomous Systems (SAsSy) project aims to enable the scrutiny of autonomous systems by allowing agents to generate plans through argument and dialogue, while justifying the purpose of each step within the joint plan. Humans or agents can then critique these plans by suggesting and justifying alternative courses of action as needed, thus driving the planning process. However, this requires the system to first present the plan that has been chosen for execution by the system to the user. Given that these plans are often long and complex, one key challenge of providing transparency to humans regarding the internal workings of an intelligent system is therefore (good) information presentation. One of the challenges has been to decide how to adapt the plan presentation to a user’s role and area of responsibility. This talk addresses a key question of how to address the “goodness” of an adaptation. It describes a layered-evaluation approach, measuring performance in a dual task effectiveness) to complement measures such as time to complete (efficiency), and self-reported measures. This talk describes two case studies evaluating visualisations of a plan that illustrates the value of the approach.

Leave a comment

Filed under Uncategorized

Videos: Adapting plan presentation, filtering and highlighting

 

One of the strands of research in SAsSy is how to adapt the presentation of plans to users depending on their interests or areas of responsibility. We have been looking at the use of highlighting and filtering for this. Our method works for both text and graphs, and considers that some tasks need to be done before others (dependencies).

I demonstrate how we do this in the current version of the system here:
screen shot of a planPart 1: Highlighting (mp4, 5.33)
Part 2: Filtering (mp4, 1.30)

Leave a comment

Filed under Uncategorized

Which forecast should I trust? Is it going to rain tomorrow?

 

What if I showed you this story?
Screen Shot 2014-06-06 at 11.33.12
The weather forecasting service of the broadcasting company AAA says
that it will rain tomorrow. Meanwhile, the forecast service of the broadcasting
company BBB says that it will be cloudy tomorrow but that it will not rain.

Screen Shot 2014-06-06 at 11.33.21

Hard to decide which one to listen to, right? (Especially when it comes to weather that is so unpredictable anyway.)

But what if I also told you that:
It is also well known that the forecasting service of BBB is more accurate than the one of AAA.

You might develop a preference for listening to BBB’s forecast. If that is the case, then you might also say that it will not rain tomorrow.

Why is this interesting? Well, argumentation theory looks at how we reason over conflicting statements. So like where AAA and BBB are saying two things that cannot both be true. The theory predicts what points of view are “logically” correct. There are different flavors of the theory though, many of which do not consider preferences.

So a theory that does not consider preferences would say that you would still be undecided even if you knew BBB was more reliable. But we felt pretty sure that people use preferences when they make up their mind about these sorts of things. And we knew of a way of representing statements like this in a logic with preferences (thanks to Henry Prakken and Giovanni Sartor).

And indeed this is what we found. People make the conclusion you would predict if they used preferences. Not only that, but they also said the preference statement was information they thought was relevant when they made up their mind.

We found this was the case for weather, but also for other domains like a car sale, and regional independence!

You can see the stories and the logical representation behind them here.

Not too surprising of course was that more people were undecided for weather than in the other domains, but they still used the preference information and decided that it was not going to rain!

My colleague Federico is going to present this paper at the European Conference of Artificial Intelligence in Prague this summer, but you can have a look at it now too.

But just in case, do not forget your umbrella!

umbrella

Leave a comment

by | June 6, 2014 · 10:41 am

Discovery day – National Science and Engineering Week.

ImageImageOn Saturday I took part in Discovery day as a science “super hero”.

I had a great time talking to kids (2-7 years old) about the “super powers” technology gives researchers. Together with other science superheros, we inspired 132 children to imagine themselves as scientists in the future (70 girls and 62 boys)!

Leave a comment

Filed under Uncategorized

Meet science superheroes!


nsew

Nava Tintarev taking part in Discovery Day at the Satrosphere Science Centre on the 22nd March.

Come along and meet some real life science super heroes, find out about their scientific super powers and make a pledge to save the world from the ‘knowledge thieves’!  Suitable for all ages, children and families.

Admission Free, no need to book. The event runs throughout the day from 10am to 4pm.

This event is part of National Science and Engineering Week – see the full programme: NSEW 2014 Aberdeen brochure

http://www.abdn.ac.uk/engage/nsew/
http://sciencegrrl.co.uk/events/event/aberdeen-super-heroes-nsew/

Leave a comment

Filed under Uncategorized

DEMOfest@RGU, 11th of February

Just a reminder that Roman will be demonstrating a new and improved version of the SAsSy system at the next DEMOfest!

When? 11th of February, 2014

Where? Aberdeen (at Robert Gordon University)

DEMOfest is a technology showcase of leading Informatics and Computer Science research from Scottish Universities and offers the opportunity for industry partners and academics to come together for:

  • collaborative innovation
  • studentships & placements
  • technology licensing & consultancy
  • feasibility & proof of concepts

Leave a comment

Filed under Uncategorized

Demo of SAsSy to be shown at the Intelligent User Interfaces conference

We are excited to announce that a demo of our system has been accepted for the Intelligent User Interfaces conference!

We will be presenting it in Haifa, Israel in February.

If you want to have a sneak peak, have a look at our video demonstrating the SAsSy system!

Leave a comment

Filed under Uncategorized

Are you talking to me? I was! At TechFest2013.

On Monday I gave a talk at TechFest titled “Are you talking to me? What to say when you are talking to Robots?”.

TechFest talk about the SAsSy project by Nava.

I spoke about what autonomous systems are and how they are different from robots. They do a lot of the same things, but
don’t have a physical form like Honda’s helper robot ASIMO, or even a Furby.
They operate independently of people, and can communicate between themselves and us people.
These systems are everywhere, in our tills, satellites, sometimes even in our copy machines (sorry,  I meant multi-functional devices!). Some of these systems are pretty powerful, like self-driving cars and drones, and so it’s really important to keep people in the loop. We need to be able to understand what is going on `under the hood’, ask why certain things are happening, and be
able to change or override system behavior when it’s appropriate.
This is where explanations like the ones we are developing in SAsSY come in – we help people understand complex plans,
and the various reasons (and counter-arguments for them, and the counter arguments for the first counter-arguments…) for
why things are done a certain way. I’ve posted the slides online, and feel free to get in touch if you have any questions about the talk.

Leave a comment

Filed under Uncategorized

Understanding WHAT is in a plan…

Hi!

Thought I’d give you an update on one of the projects I’ve been working on.
One of the high level questions we’re trying to answer is how to present a sequence of actions or a plan. We realized that a person needs to understand WHAT they are supposed to do before they start asking questions about WHY the plan is the way it is.

So, for the last couple of months I’ve been setting up and running experiments on mechanical turk (fondly called mTurk).
MTurk is a crowdsourcing Internet marketplace that enables individuals or businesses (known as Requesters) to co-ordinate the use of human intelligence to perform tasks that computers are currently unable to do.

Best performing text plan
Best performing text plan
One of the interactive plans
One of the interactive plans

The first experiment looked at six ways of presenting a pizza recipe. It had pineapple and banana on it – you can imagine the sort of comments we got on our topping choices! Anyway, we compared three textual plans with different levels of headers and slightly different formatting, and three interactive plans.

We figured that some ways of presenting the plans would help more for answering questions about the recipe, and we measured cognitive load (roughly mental effort) in a few different ways. Surprisingly, one of the textual versions tended to do better!

So we ran another similar experiment with larger plans (125 steps) – perhaps the advantage for interactive plans will come through when there is too much information to take in at once we thought. These plans were about delivering things (televisions, screens, laptops etc) to different locations.

We also looked at using aggregation, or a label like ‘move the object’ to summarize three other actions: load a truck, drive the truck and unload the truck.

We are currently in the process of analysing the results of this second experiment!

Leave a comment

Filed under Uncategorized