About Me

Deana Bui

Mills College, Computer Science

Senior, Graduating 2018


Contact Me

deanabui@gmail.com


Journals

07082017 - Tenth Week - Last Week, Last Chances

The program participated in a symposium where we all felt slightly awkward as it was one for natural sciences, but we enjoyed it nonetheless. We even had the laptops out to demo what the operators do and see and how they would interact with the users. Some of the staff and other students from our lab even came to see and support us.

It was rough adapting to Minneapolis and trying to get settled down while doing things full force. Rough, but honestly was a great experience now that I look back on it. I met some really great people and made some really great memories.

Although we weren't really able to kick-off our experiment. I'm leaving feeling confident that it will be able to. I'll be keeping in touch with my group mates and participating however I can.

31072017 - Ninth Week

We're cutting it close with the IRB approval. We initially got rejected, but with some minor fixes we should be good to go.

UPDATE: WE ARE GOOD TO GO

The operator interface is a game changer for how we operate things. We've set phases to timers and things are sent automatically now. It's possible to run tests with one person with these additions.

We're seeing some patterns with how users are interacting with the bot and how receptive they are to it. There's disdain towards bot behavior that is restrictive with user actions. Users generally become frustrated with the bot and end up defaulting to a movie that is popular and well-liked (i.e. Disney, "classics", superhero movies, etc.) Users, howevers, do like the bot for the reason that it is helping them explore options that have never heard of before.

There's only one week left. I wasn't expecting to actually start the testings given that the IRB application process was new to all of us on board. Besides all that, it was a great learning experience.

24072017 - Eigth Week

Finally submitted our IRB!!!!!!!

We also got two more pilots tests in. One for control and another for the active structure. We're not quite sure what to do with the results we got since it's kinda an outlier case. Also had some moments of realization wondering if this study will even hold. On a brighter note the second interface works wonders in our timing. If we really wanted to the bot operation would only really need one person, it'd be chaotic, but doable.
We're hoping it does and it's just some details that need to be adjusted and refined. We're confident.

UPDATE: I pushed for a test with the passive bot and the results were definitely what we needed to see. It reaffirmed what we were doing and with some changes this could happen. We're planning for another pilot test or two next week to see how that goes. Hopefully we could get this up and running before Mary and I have to leave Minneapolis.

17072017 - Seventh Week

We had our first two pilot tests this week and whhoooo were things janky. It was difficult for Mary and I to deal with the bot script during discussion. We're talking of creating another interface for the second operator solely for dealing with dialogue. It was stressful on both ends for our team and the participants. Stressful, but definitely got some feedback that we needed.

We have so many copies of our bot scripts. So Many. I don't even know which number revision we're on anymore. It's difficult to pinpoint where we're out right now.

Our inside joke of the week:
Person: What did you do for your internship this summer?
Us: We wrote tons of scripts!
Person: Oh cool! In javascript? Py-
Us: In English.

10072017 - Sixth Week

Went to a peer review and received some good feedback and concerns about the project. We've answered some questions we have in regards to our recruitment pool and research questions.

Also Mary and I have adopted a fungus named Finn "the Funguy" Fungus.

03072017 - Fifth Week

Created the first version of our script for the chatbot and drafted up potential questions we would ask in a feedback survey.

26062017 - Fourth Week

Finalized more details this week. Changing directions after the group meeting. There are worries of if we're creating biases with the bot dialogue. Josh is also implmenting an API that will allow us to pull information easily rather than needing to do each individual detail manually. This would solve some slow timing issues.

We're leaning towards recruiting from MovieLens & community though this might be an issue if we're looking towards scalability, but this would sacrifice the validity of the data collected.

19062017 - Third Week

Still reading those articles. We started drafting our ideas down and figuring out the details of the study. We're not quite sure which direction we're going in yet, but I think we're off to an okay start.We've also created a 'script/dialogue' of the chatbot responses.

We had a small meeting with Max to go over details and we got some good feedback. Got a good idea of adding a voting poll to the active bot. We're now deciding which chat platform we want to use where the bot will be interacting with the users, intelligent agents, and where we're going to recruit our users from. Thinking about Amazon Mechanical Turk, existing MovieLens users, or from our own communities.

Plenty of work to do.

12062017 - Second Week & The Beginnings

Met Frank, a Phd. student working on the user study portion of the research. We discussed and broke down which project we were working on and what we were researching on.

Mary and I worked on the training that was required for researchers engaging in human subjects research and read some articles. There were plenty of articles on bots, but I found the language aspect of the research more interesting. I found a 400 page book on the literature of group-decision making.

It was raining this weekend so we didn't plan any outdoor activities. We decided on going to the campus art museum. Hands down the interactive apartment installation on eaves dropping was my favorite.


  • Finally got our student accounts/IDs.
  • Finally got access to the secure wifi connection.

04062017 - Move In & First Week

Moved into the dorms in the evening. Ate some pretty good Himalayan food. Met the rest of the DREU (DREU-CREU grandmas) and Big Data REU members on Tuesday. Met Haiyi and the rest of the the project group on Wednesday. We discussed the potential projects that we can work on but nothing is set in stone. Mary and I have to wait until Monday to find out what we can actually work on, but Haiyi sent us some articles so we can get some idea of what user study research entails.

Went to the Mall of American with some of the REU. There were Annie auditions going on. Looks like the sun is coming out tomorrow.


  • Using the guest wifi is a struggle.
  • We still don't have our student accounts/IDs.
  • The Mall of America is still as big as ever.
  • Lake update: There's more than 10,000 lakes.

01062017 - First Time in Minnesota & Housing Struggle

The moment I touched down I received a message from the person I'm suppose to be renting from saying that they had to cancel because their friend needed a place to stay. Great. Entirely greatful to DantÄ— for letting me stay with her for a few days.


  • Housing Struggle cont. - Staying at the dorms instead.
  • Yes, the Mall of America is big.
  • They weren't kidding when they say it's the land of 10,000 lakes

MovieLens Bot

Group Decision Making

Defining the behavior of the bot

    Active

    • Actived bot facilitates the discussion process in addition to carrying out the services of the passive bot. The following bot designs make up the two interventions of the active bot.

    Passive

    • Passive bot assists their groups by providing movie information or movie recommendations when directly requested from the users.

    Opinion

    • Opinionated bot gives comments on movies, based on the user's preferences (indicated in the MovieLens signup). For example, it will point out whether there are positive or negative opinions and whether the movie being discussed will be a good or bad match based on their group preferences.

    Structure

    • Structured bot enforces the three phases of the decision making process; Brainstorming, Discussion, and Decision.

The Mentors

Hao Fei Cheng

PhD Student

University of Minnesota, Computer Science

Human Computer Interaction, GroupLens


Frank's Website

Maxwell Harper

Research Scientist

University of Minnesota, Computer Science

Human Computer Interaction, GroupLens


Max's Website

Dr. Haiyi Zhu

Assistant Professor

University of Minnesota, Computer Science

Human Computer Interaction, GroupLens


Haiyi's Website

The Labmates (Struggle Buds)

Mary Solomon

Co-Secretary, The Researcher

Bowling Green State University, Data Science


Josh Wissbroeker

Operator

University of Minnesota, Computer Science

LinkedIn

Elements

Text

This is bold and this is strong. This is italic and this is emphasized. This is superscript text and this is subscript text. This is underlined and this is code: for (;;) { ... }. Finally, this is a link.


Heading Level 2

Heading Level 3

Heading Level 4

Heading Level 5
Heading Level 6

Blockquote

Fringilla nisl. Donec accumsan interdum nisi, quis tincidunt felis sagittis eget tempus euismod. Vestibulum ante ipsum primis in faucibus vestibulum. Blandit adipiscing eu felis iaculis volutpat ac adipiscing accumsan faucibus. Vestibulum ante ipsum primis in faucibus lorem ipsum dolor sit amet nullam adipiscing eu felis.

Preformatted

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';

Lists

Unordered

  • Dolor pulvinar etiam.
  • Sagittis adipiscing.
  • Felis enim feugiat.

Alternate

  • Dolor pulvinar etiam.
  • Sagittis adipiscing.
  • Felis enim feugiat.

Ordered

  1. Dolor pulvinar etiam.
  2. Etiam vel felis viverra.
  3. Felis enim feugiat.
  4. Dolor pulvinar etiam.
  5. Etiam vel felis lorem.
  6. Felis enim et feugiat.

Icons

Actions

Table

Default

Name Description Price
Item One Ante turpis integer aliquet porttitor. 29.99
Item Two Vis ac commodo adipiscing arcu aliquet. 19.99
Item Three Morbi faucibus arcu accumsan lorem. 29.99
Item Four Vitae integer tempus condimentum. 19.99
Item Five Ante turpis integer aliquet porttitor. 29.99
100.00

Alternate

Name Description Price
Item One Ante turpis integer aliquet porttitor. 29.99
Item Two Vis ac commodo adipiscing arcu aliquet. 19.99
Item Three Morbi faucibus arcu accumsan lorem. 29.99
Item Four Vitae integer tempus condimentum. 19.99
Item Five Ante turpis integer aliquet porttitor. 29.99
100.00

Buttons

  • Disabled
  • Disabled

Form