Huge Interface Revision Coming Up Before Final Presentation.

affectiveflickr. 1.3 Version


Basically it involve a huge implementation improvement.
That's the social collaborative  filter.

Collaborative filtering

If user A input  "happy" in the website, then some picture came up.

e.g. Picure1  

User A select scores 9 for this picture then saved, he press next

Picture 2 came up

User A select scores 4 for this picture then saved.

User B came he input the exact same tag as User A which is  then Picture 1 will show up first because Scores 9 > Scores 4.

So generally, picture with higher score will shows firstly then picture with lower scores will shows lately.

Conditions: Tag must be the same.

So database should able to store the every tag

Interface Changes:

Add a small bar under the picture windows shows that more than one result of the picture, for example. If input  something then it will shows several result in once. Therefore user could see some other pictures as well.

Other Changes:

Numbers of small bug has been fixed.

more stable and reliable voting system.

new improved database methodology.

Faster porting time.

Draft Treatise

The Draft Treatise will around 5000 words, and cover most of my work but without the result analysis and the final conclusion. I will list the table of content below which will give a brief go through of my draft treatise.

Table of Contents

Acknowledgments. 1

Abstract. 3

1 Introduction. 4

1.1 Why are emotions important. 4

1.2 Can machines ‘feel’ emotion?. 5

1.3 Developing Emotions. 6

1.4 Basic Emotions. 6

The Nature of Emotions. 8

2 Background. 10

2.1 SAM – Self Assessment Manikin. 10

2.2 IAPS-International Affective Picture System.. 13

2.3 Flickr®. 15

3 Development (Draft) 17

3.1 Emotion Input. 17

3.2 Collect Affective Data. 18

3.3 Emotion Output (Unfinished experiment) 20

Bibliography. 23

Possible implementations to improvement the user interface

After a short disscussion with Dr.Raefal I have few idea come up which can improve the user interface to make the website easier to use and understand.

There is a way to simplify the voting process is to combine the axis and the emotion icon panel, the horizontal axis which represent the Pleasureness, and Vertical axis is the Arousal.

Pleasureness

Arousal

so combining the above diagram we get the following

User only have to drag two bar to vote which can delete the axis panel, however the axis panel still will appear on the website just no longer to point it.

SAM analysis

The IAPS consists of hundreds of pictures created as ‘‘a large set of standardized, emotionally-evocative,internationally-accessible,color photographs that includes images across a wide range of semantic categories’’ (Lang et al., 1997, p. 1). One goal of the developers of the IAPS was to provide a standard stimulus set that could be used across a wide variety of studies on affect and emotion.This goal has been very successfully met, with IAPS pictures being used in hundreds of studies in the behavioral, social, and
neurosciences.
One reason for the popularity of the IAPS is that each picture has a normative score for males, for females, and for both sexes combined on the Self-Assessment Manikin (SAM). The SAM was introduced by Lang (1980) as a quick, nonverbal method of quantifying subjective feeling states on the three fundamental dimensions of evaluation identified using the semantic differential in work by Osgood and associates (Osgood, Suci, & Tannenbaum, 1957) and by Mehrabian and Russell (1974). The SAM was created to reduce the language requirement of the scale so that children could provide ratings using the same instrument as adults and so that ratings from different cultures could be compared. The SAM has gone through several iterations in both computer-administered and paper-andpencil formats, but its basic characteristics have remained unchanged. A series of graphical figures are used to elicit a single rating for valence (ranging from extremely pleasant to extremely unpleasant),arousal (ranging from extremely aroused to extremely calm), and dominance (ranging from totally controlled to totally in control) dimensions.

These three aspects are not parallel to each other, actually they can form a three dimension phase.

Website Database Done

After the debugging process, my database for the website finally finished.
Here I want to talk about the database and give a brief explaination to the database.

Above Image is a screen shot of my database, there are four seperate tab for this database: images, picNum, rating and scores.

Images Function : Record the URL address for a certain picture viewed by user.For instance so you will  see a address like this : http://farm5.static.flickr.com/4059/4475929936_a88c5f442c.jpg. and a id number.

Score :
ID-Will distribute a integer value to the access order
p_id-Is the integer values for a certain picture,which link to the Images, so we could track which picture has been voted.
type-Is a boolean values which only have 1 and 2 , "1" mean the emotion icon votes, "2" mean the axis panel vote.
score for "2"- This value comes from the emotion axis, which is a integer value arrange by the file point,
point is a flash object in the website ( I will give a detail discussion on future blog post). there are about 7000 points on the axis panel, as the following images shows.

score for "1" - This value is a average value, there are seven image icon on the website so. from 1 to 7,

if user A press the first one, user B  press the fifth one so the value for "1" is equal to (1+5)/2 = 3

Hopefully this post will give a better understanding to my database record process.

http://www.affectiveflickr.com/ is now online.

Further Application About SAM

Understanding emotions is vital to understanding people. Every decision is a combination of rational and emotional factors. And emotion has been shown to be the driver of behavior and preference.

So Why are emotions important?

Well,emotions is a powerful influencers which effect behavior, relationships, evaluation & consideration, to name a few.

Adsam is a assessment model which specialised applied to advertisement, brand and any other form of design, adsam is a non-verbal measure, which eliminates verbal biases and makes its appropriate for cross-cultural use. The globally recognized system includes measurement, analysis, consultation, written reports and recommendations. As we know SAM ( Self-Assessment Manikin) can analysis three aspect of emotion. Pleasure, Arousal, and Dominance, so the user of Adsam can get the result from those aspect and decide which of their advertisment design is more "Emotional".

New Improved Interface

The New Interface for the affectiveflickr has been done. It has the following Improvement.

1.More Friendly Interface which will give user more directly view the picture.
2.The picture now has the automatic fit to the layer. so the picture can be any resolution.
3.More flexible emotion evaluations. User now can point any coordinate in the axis panel,(before is fix point selection).
4. Drag function, User now can drug the emotion icon to the right hand panel.
5. History Function, user now can be review their or other's old submission on the picture.
6. Output file, we can get a new output file which anaylsis all the emotion submission data on a single picture.

Will Upload the Completed Version to the domain : www.affectiveflickr.com