#8 Progress report

Using ManyEyes 
This week I have been working on my 3-5 minute presentation and I have also started working on my next step, which will be to look at five press releases from BP’s homepage between the dates 21 of april to 28 of april and two articles. As I wrote in my previous progress report, these articles and press releases are then put through IBM's text analisis program ManyEyes.

I have started looking at the articles and press releases with ManyEyes, but unfortunately I haven't had the time to start working on this, so my project have pretty much stood still this week. 

So I will update within a few days...

#7 Progress report

Creating my presentation platform
First of, I've created a google site, using on of the existing templates, which was thought to be my medium of presentation of the project. Then I modified the template for a more user friendly environment. Sadly, all of this turned out bad. I erased to much of the sidebars and lost track of the red line of the site. So, instead, I created a separate blog, here on blogger: Final Project

The header consists of pictures of the oil spill in the Mexican Golf, I’ve photoshoped different pictures and created a montage which captures the oil spill disaster in a broad perspective. I have also included BP's logo in one of the squares, in this way, it makes it more understandable of which oil spill I have created the interactive map. I'm not sure thou about the copy right rules of the pictures in the header, and I need to look into that and see if I can use them...


A interactive map of the Gulf oil spill

The interactive map includes the position of Deepwater Horizon, the fishery closure boundary and also the position of the Gulf Islands National Seashore which is a protected territory because of it’s rich wild life. I’ve found a problem when using the line tool. I tried to insert a fish symbol for the fishing boundary line, but with out any success. I e-mailed google about this and we’ll see if I can get any suggestion on how to solve this. I think it would be more sense-making to have a symbol somewhere along the line to make it more understandable of what just that specific line means.

Text analysis

I have picked out five press releases from BP's 'Gulf of Mexico response' web page, they are all within the first week of the disaster. And I have also picked out some articles, just to check if my analysis program ManyEyes can create interesting visualizations. I'm pleased with the results and I believe ManyEyes is a suitable tool for my material. 

Problems and thoughts

1) Find a interesting approach on my material. I have still not found a approach for my material, and I need that before I can start my analysis.

2) I have decided that I'm only gonna create a interactive map, instead of both a map and a timeline. I think it will be enough, if I can place findings of my analys and keep on building it up with more tags with relevant facts.

To see the Final project, click here.

Remediating History through Mapping

I picked the battle of Dieppe because World War II is interesting in many ways. It has shaped and influenced the world as we know it today, both culture vice and geographically. All the information used in the map is taken from Wikipedia. I have selected parts from the event which I think works well with the look of the map.

My remediated map consists of three different tags. One representing a Nazi German radio station, which was one of the targets of the mission and one image representing a tank regiment and finally a tag which indicates where the army was put on shore.
I used clip-art images to capture and symbolize the action, in this way it makes it more interesting and also more understandable. I wouldn't go all the way and say that I create a realistic visualization, but something in that direction. In the tags you can find short information and a picture.
I have included links in the tags which takes you to Wikipedia for further information. As for the metadata, I chose to show the source from where I got them. I wanted to include a short movie clip of the event to give a even more clear picture of what happens, but could not find that option.

History Pin is a awesome tool for students whom doing history documentation projects. It is user generated, as for both pictures and stories and collaborates with google street view. This means that while you are using google street view you can also see how a building used to look etc.
Community Walk, differs however from History Pin, it is mainly a site for mapping communities and building interactive maps. Its target group are business and property owners whom wants to market houses and districts. Trippermaps is a website which allows you to make photo collaborations with your Flickr pictures. The target group is mainly blogger whom can use this to show of their photos in a more interesting way. Trippermap, compared to History Pin and Community walk, is not built on Googles maps interface. Insted it uses its own platform.



My creation of the 1942 battle of Dieppe Map

Final project

For my final project, I will look closer at the disastrous oil spill in the Mexican golf. This disaster occurd the 20th of april 2010 and lasted for months. It had a massive impact on humans as well as the nature, and it will take time for the natural life as well as for the humans to get back to the life and surroundings as they know it. Therefore, I think it can be interesting to to use the features as ManyEyes to perform text analysis on press releases from BP during the disaster. Due to the limitation of time, I will not be able to go thru them all. Therefore, I need make a selection. I want to see how they communicate in the beginning, middle and in the end of the catastrophe. ManyEyes allows to find similarities and patterns. I will also create a Timeline. Which, hopefully, connects with my text analysis, in that way that I can see patterns.

The press releases can be here 
http://www.bp.com/PressReleaseDateSorterAction.do?categoryId=2012968&contentId=2006635&year=2010&month=4&taxonomyId=2021012

This will be my last blog post on this blog, my work will be shown on a separate blog.  

Visualization of The Count of Monte Cristo

Plot.
If you are not aware of the story here comes a short summarize: A young and successful merchant sailor (Edmond Dantes) returns to Marseilles to marry his love, Mercedes. Shortly, a anonymous letter accuses Dantes of being a Bonapartist. He is wrongly send to jail, and the rest of the story is about Dantes revange and his love to Mercedes. Danglars, which can be seen in the wordl is the one who sets Dantes up. The other one is Villefort who also is visualized rather large in the wordle. 

• Title / Name: The Count of Monte Cristo
• Author / Creator: Alexandre Peré Dumas
• Publisher: N/A
• Date of Creation: 8 November 2008
• Date viewed: 18 October 2010
• Persistent Identifier: http://www.gutenberg.org/files/1184/1184-h/1184-h.htm
• Language: English
• Format: .html
• Media type (if applicable): N/A
• Subject or Topic:
• Category or Categories:
• Tags (if none are available, write down the tags that you would add): Romance, Revange, 1800th story, Dantes, Chateau d'if, Monte Cristo, Alexandre Dumas

First off, I'm gonna use Wordl to make a visualization of the text.

Settings.
Most frequent word used: The
Content word: One
Layout: Straighter edges and Half &Half
Color: Moss & A Little Variation
Font: Sexsmith
Words displayed: 200

Wordle gives me a visualized overlook of the text, at first look many of the main characters is shown, but if look closer you can also tell that the language is not only English. So, for a person who has not read the text, they can conclude a French touch of the story. The words that is visualized are locations, the characters names, and the story in short single words. So it summarize the text pretty good, but I think Wordls plattform, or the program is limited. I would like a feature were you can ask more questions from the text, and maybe put two and two together and create text analyzes from your created material. 

Text analysis with TAPoR.
TAPor is a text analisis program. This program is perfect if you want to create metadata about your text that your working with. The text can be of any kind. I would say that this program is suited for people who has large texts and for some reason wants to know specific information about the text. Say that you would want to know how many palindromes there are in the text, palindromes are words that are spelled the same weather they're read forwards or backwards. This would probably take for ages, and it's easy to make a mistake, and then you would have to start all over again.

As can be seen from my screen-shot, the possibilities are endless on how you can analyze a text with the TAPoR program. You can also find hidden patterns with this program.

Beneath are some metadata of the whole book, TAPoR creates this kind of data super fast, and it doesn't just look for simple metadata as page numbers or amount of titles etc. It can give you the average number of words used in a sentence, unique words, average word length etc. These facts might not be of any value to the common reader, but it is actually valuable information if you are looking for similarities or word concordance etc.    

 

Visualization with ManyEyes.

Here is my text presented in the IBM software ManyEyes, which is a text-analyze program with visualization features. You can choose to view your text in different visualization settings, the one beneath is the word tree. It's function is to show you how a word or a phrase is used in the text. For my text, Dantes is the main character and this is the findings with only his name.

  It doesn't just stop here, I can go further in my investigation. If I then click on another word, which in this case is 'was', it gives me all the phrases with 'Dantes was'. This tool is valuable, because like the other analyzing tools it gives you another perspective of the text.

And you can continue clicking on the words provided, until there is no phrase with the words you chooses. 

Summarize.
To analyze and figure out how we can do things better is in human nature, and these tools shown above are helpful for analyzing texts, but it has not yet spread the the broad public.
Companies uses analyzing tools daily to create better process and uses it to filtrate among information, just to be more effective and create understanding of what happens in different scenarios.

But, the tools shown, are not just for companies, they will eventually become "googelised" to the broad public and with that we can filtrate information even better then we can today with our search engines. As a user, it is not the word you put in that you are after, it is the meaning your after. And that is what I think these tools might give us in the future, but as the work today some of these tools are rather user friendly, but the creations it gives us might not be needed for the broad public.


As a student, these programs can come in handy if you need to get perspectives of your own text, or you are looking for clues on how to criticize some one others essay. ManyEyes different ways of presenting a text; scatterplot, bubble chart, pie chart, treemap etc are excellent ways of presenting rough data, that could be hard to explain in a short period of time or just to rough for common people in general. So I would say that these programs are best suited for teachers (in any level) students (in any level) and for people with the need for visualization tools at work, and that can be quit many!

Blog assignment #4. Visual communication with Visualization lab.

Today we worked with New York Times Visualization lab, which is a new text analysis feature built on IBM’s Many Eyes. This feature makes it possible for the user to create visual representations of information and data, like maps, charts and other graphic presentations out of the articles posted on NY Times homepage.

When does this come in in handy, and why do we need it?

This is a new way to interact with their online users. It can come in handy especially with texts that involves politics, where the usage of words can give away a lot of the content of the text. The charts you can make can put other perspectives on the text, it can help you form questions on the subject and clarify the sours standpoint. So, the tool can be of help in many ways. Charts and such has been around for ages, just because it is a simple way to represent statistics and facts.

• What kinds of data can you analyze using your suite of tools?

NY Times articles.

• What kind of information can you extract from the data?

You can create visual representations of the articles.

• What kinds of questions can you ask of your data using text analysis and data visualizations?

The main thing with this tool is to create discussions. You can analyze the text and make statements in the commentary area.

• What hidden patterns are revealed using text analysis and data visualization?

Well, the patterns are there, maybe its not the pattern that you thought it would be, but this is a tool to find new information about the information…

• Who would be most likely to perform this kind of text analysis or data visualization?

The point is that every body can use it. It is a way of integrating this into articles to make more discussions out of it. 

My Text analysis and data visualization exercise. I'm gonna look at the novel: The Count of Monte Cristo. I will see if can come up with some good metadata and visualizations for the text, using Wordl. Many Eyes or TAPoR, which is other text analyzing programs might also be used... 

Creating word clouds

Text 1. What’s the Point of a College Education?

No changes made with the text.

Costumizations: None
Color: yarmirP
Layout: Straighter edges & Any Which Way

This text seems to be about collage and education which adds up with the title of the text. It is hard to tell if this text is positive or not based on the words, you need the context of the words to figure that out for sure. I'm not surprised by the words that appears, it goes well with the title, which gives it all away. If I take away the word 'Collage' the cloud changes, and the most frequently used word would then be 'Education'.

Text 2. Blog assignment #2
Font: Duality
Layout: Straighter edges & Mostly Horizontal
Removed words: 'Tags', 'Video' & 'Specific'

Clearly this text is about metadata and information, and those two words goes hand in hand. I had given the text title 'Blog assignment #2' which doesn't give much information of what the text is actually about. Here is a perfect example of what 'word cloud' or 'wordle' can help you with. If I would take away the word 'information' which is the most frequently used word the cloud would change, and it would even out the size difference between the words, and that means that the words are used as much.

Text 3. George Orwell’s 1984
Font: Duality
Color: WB & A Little Variation
Layout: Rounder Edges & Half and Half
Removed words: 'Winston' (The most frequently used) and 'One'

I haven't read this text at all and I have a hard time figuring out what it is about. The most frequently used words 'Party', 'Even', 'time' etc doesn't give me the context. But it is a positive text, based on the cloud. Or, I can't see any proof that this text would be a negative one.

I took out the most frequently used word: "Winston' which is a name, and I guess the text is about him. But the word cloud above is with out his name, and that makes the cloud harder to understand.

Text 4. Transcript of Commencement Speech at Stanford given by Steve Jobs

Font: Expressway Free
Color: Milk Paints

Layout: Straighter Edges

Removed words: 'mother', 'began', 'next', 'got' and stay.

Steve Jobs is famous person for what he has done with Apple, and my first thought was that the word cloud would be about Apple, technique, Macintosh etc. I deleted some smaller words in my wordle so the most frequent words still would be kept.
The text is a speech given at a University and it seems to be about collage and life, and maybe how Apple sees it or something. The word 'death', which is in the middle seems a bit negative compared to the other words around it. The words I took out didn't change the cloud in any specific meaning, which was my point for this cloud, I wanted to erase some smaller words to see if the changes, and I wouldn't say that there was any dramatic changes.

Text 5. A Poem from Representative Poem Online

Font: Scheherazade
Color: Costume Palette, A Little Variation
Layout: Straighter Edges, Any Which Way
Removed words: 'Well' (most frequently used, 'first' and 'thou'

I would say that these are words that you would find in a poem. Two words can be seen which symbolizes love; 'lovest' and 'lov'st'. These words are positive, so we could assume that this poem is about love. I took away the most frequent used word, 'well' because it can mean different things and really doesn't say much about what the text is about. The cloud changed and now the words about love is the first you see. I think it symbolizes the poem better with these words. 


D.  Now it’s time for some fun! Create 3 more wordles of texts of your choice:


6. Wordle of my facebook profile.

Most frequent used word: Johannes (39)
Content word: 'Remove'
Removed words: 'No' and 'View'
Layout: Rounder edges & Horizontal
Color: Moss
Font: Owned

I removed the words 'No' and 'View', these words doesn't mean anything in this context, the cloud changed and my name is now the largest word= most frequent used. I'm not surprised by the words thats in my cloud, this is actually the words that I would assume would turn up due to the fact that it is my profile. 

7. A critique essay of a drawing.

Removed words: 'Windows', 'Just' and 'One'
Color: Organic carrot
Layout: Straighter edges & Any Which Way
Font: Sexsmith
Most freaquent word: 'The'
Content word: 'Drawing'

The essay is 5 pages and I used the word 'The' 216 times, I assume this is because you use it to conjugate words in English. The essay is a critique of a drawing, and thats my content word. I removed the word 'Windows' which was my largest word presented in the wordle. I think my word cloud gives a good picture of what my drawing presented; a building with different actions in all the windows of the house.

8. Owners of Red Sox to Buy Liverpool Soccer Club

Color: Custom palette
Removes words: 'Venture' and 'Times'
Layout: Rounder edges and horizontal
Font: Grilled Cheese BTN
Most frequently used word: 'The'
Content word: 'Owners'
Link to text: http://www.nytimes.com/2010/10/07/sports/soccer/07liverpool.html?_r=1&ref=soccer

The article is about changing owners in a famous European football club. The wordle symbolize the article good, it actually gives you a lot of the information provided in the article. I cant answer if the words are positive or not, you need the context to draw such a analyze. I removed words which are not needed to understand what the text is about, it makes more sense now.  

9.

Is this tag cloud different than the wordle you created?
It has not changed much, the message remains the same. You can understand that there is something about the owners of the football club.

What options does Tag Crowd offer that wordle does not?
TagCrowd is taking tag clouds far beyond their original function:

• as topic summaries for speeches and written works
• as blog tool or website analysis for search engine optimization (SEO)
• for visual analysis of survey data
• as brand clouds that let companies see how they are perceived by the world
• for data mining a text corpus
• for helping writers and students reflect on their work
• as name tags for conferences, cocktail parties or wherever new collaborations start
• as resumes in a single glance
• as visual poetry

Sours: http://www.tagcrowd.com/blog/about/

What other features would you like to see available in Wordle or Tag Crowd, if any?
I would like to have these features in programs like word and excel, or in my phone. It would make analysis of texts easier.

What is a stoplist?
A stoplist is a list of stopwords that, when associated with a full-text index, is applied to full-text queries on that index.