What Are Your Research Ideas?
The idea for Media Cloud emerged after a series discussions between faculty and friends of the Berkman Center. The conversations would follow a predictable pattern: one person would ask a provocative question about what was happening in the media landscape, someone else would suggest interesting follow-on inquiries, and everyone would realize that a good answer would require some real number crunching. Nobody had the time to develop a huge infrastructure and download all the news just to answer their one question. However, eventually there were enough of these questions that we decided to build a tool for everyone to use.
So what are your ideas? What questions could Media Cloud help you answer? Leave them in the comments below, and we’ll talk about them together. It will also help us to build the system to support your ideas.
I’d like to know what media sources from other countries are saying about the US presidential inauguration.
This is great!
I’d like to know how media scandals break around the criminal justice system, and whether there’s an independent role for the progressive/black blogospheres in getting these stories out.
My question is: are you tracking stories retroactively, or starting from now on? I have a big question that I want to ask that would require me to start in 2006.
What a great idea! Part of the problem with media has been the difficulty for the viewer public to gain a sense of objective information by balancing delivery perspectives and, indeed, the focus of overall coverage. From my very limited understanding of the Media Cloud concept, viewers will at least now be able to see where media resources being concentrated – another element of the ‘balance’ idea.
Using media resources to engage in competing views about the argument over a neighbor’s dog means those resources are not available to cover the re-zoning board meeting which could turn the whole neighborhood into a commercial zone – or the proposed federal regulations which make the small town’s main street a new highway. Just to provide some simple (entirely fabricated) examples….
I’d like to know what media sources are saying about the FDA and pharmaceutical risk management.
Love the idea – what’s the chance of it becoming multi-lingual? For example, as a media teacher in Norway I see the possibility of using a tool like this to demonstrate how Norwegian and world media cover the same event.
I’d like to see media mentions of health tech and IT (per the current debate) – would like to use NLP, semantics, ontological processing to determine usage of terms EHR (electronic health records), EMR (electronic medical records) and PHR (personal health recordst).
Document the progressive thaw in coverage of Israel
Please work in a “velocity” tracking ability. I’d love to see measures of information movement like: where and when sources of information occur (including increases in velocity as more sources pick the story up); how the story broadens to include similar events in other locations (i.e. added local flavor or connections to historical events); a cartographic mapping of actual locations (as well as can be determined) of the sources and subjects of stories to see how “local” news becomes global and get a sense of how news is altered by the global game of “telephone” as news migrates.
In addition, I would love to see historical background information linked in. If tension is rising between X and Y, what changed recently? What history lies behind this change? Something like a timeline subtext that shows how multiple variables changed over time so that you can see developing patterns that are impacting this current situation – or that you can start to see where a conflict may be headed.
Stories are never truly two dimensional as they are often portrayed in the media (black and white based solely on current “facts”). There is huge multi-dimensional context behind everything that happens. The more we can pull that all together and display it in a rich tapestry of information, the more we can stretch our minds and better grasp the complexity of the world around us in a quicker more intuitive way.
It’s time consuming to read thousands of different articles from hundreds of different publications, encyclopedias, blogs, etc. Process it, distill it, link it, don’t hide or remove it. Embed it all in the tapestry in condensed, woven form.
I’d like to know about the links between different content dealing with energy topics. Where do stories start (date/time), how do they propagate (order by date/time) and through what routes (who links to who) ?
This should allow me to find content originators and carry out a step by step analysis of if, and how, information quality is retained as content is propagated ?
Quantitative Chinese Whispers, if you like.
I would support what Ian Falconer has proposed. Being able to study the original moment of a piece of information and then track its circulation, seeing how it perhaps alters in its travel, would prove invaluable in understanding the workings of the intertubes.
As a psychologist, I’d like to know how people are coping with the economic crisis. Are they meeting with friends more often? At home, as opposed to in bars or restaurants? How are their social lives affected? Thanks!
Thank you for undertaking such an ambitious and valuable project. Here are a few thoughts:
We need to be able to compare many more sources at a time, and to group them for a category comparison (financial news sources vs general, for example.) Any geographic results should give us a downloadable KML file that can be opened in Google Earth, or, better still, perhaps, a button that says “View in Google Earth.” Ideally, any time series results should be equally exportable as CSV or other standard format.
Next, imagine you wanted to watch the spread of financial news and map it against market behavior; just as the better financial services allow us to superimpose several variables, so might your service permit us to show activity related to the stories, but from non-news sources such as market activity, disease outbreaks, occurrences of violence, etc.
Finally, and I’m probably dreaming here, it should be possible to run some statistical tests on the results we get, right within your software.
BG
[All very good ideas. We hope we can do them at some point. We'll also be releasing the source code, so maybe you can help! -Ed.]
Most of events, actions are linked to some source of news or news article; if you could map them, it can be an great tool for predictions as well pattern understanding.
I’d like to know how “blackness” is being discussed in European countries with growing populations of people of African and Caribbean descent as compared to the US post-Obama.
This has such enormous potential! Thank you. I’d like to see an FAQ for the site eventually. A link to a list/database of the media sources would be useful; if I want to suggest sources for you to include, it would be helpful to see if they are already on the list.
Because I want an international perspective, I hope your database will start including some of the major papers from other countries. Le Monde Diplomatique, La Jornada (Mexico), El País (Spain), are just a few examples.
Are there other similar centers in other countries that would be interested in partnering to build the tool in their respective languages? A Spanish-language version would already incorporate a few dozen nations.
For a given country or region, identify the media sources that provide the most coverage.
I’d like to track the impact of the educational goals set out by the president, particularly as it relates to use of stimulus funds, state actions including legislation on educational reform, and measurable outcomes from educational institutions
I’d like for you to include “Democracy Now” (http://www.democracynow.org) among the news sources listed.
Are you planning to design an algorithm that can retrieve classic examples of slanted coverage, such as phrases like “so and so was TRYING to do x,” or “some people say . . .” Also, I’d be interested in finding some way to document the quality of a translator’s voice (translators employed by particular media sources), and how that tone of voice/ability to articulate the language clearly, may or may not relate to a particular slant on an issue.
I’m interested in coverage of ‘Octomom’ and how this could be related to discussion of health care coverage.
I would love to see the ratio of news to commentary about the news. Certainly a large amount of cable and blog coverage is pundit commentary, but it would be good to know how far media as a whole is devoted to navel gazing.
This project is awesome! I look forward to following your progress. I am a teacher educator and believe this could be a very useful tool for identifying effective teaching practices and trends and issues in education. Particularly, I am interested in success stories and challenges surrounding the inclusion of students with disabilities in K-12 schools and postsecondary institutions.
Best,
Sarah
I’m looking for information on health care cost control in developing nations and human capital reporting in listed firms of developed and developing nations. Would there be any chance for the following:
1.) Would it be possible to support Special Interest Groups (SIGs) comprising of interested persons who would like to pursue a particular line of research (eg. interested academics, corporations (including indiviudal employees of corporations, members of the civil service, etc.)
2.) SIGs comprising a wide range of participants would be able to provide coverage of the nations less covered by the given maps (as indicated in the comments so far).
3.) Would be happy to follow through if there is interest in this idea.
I’d like to know what sources people turn for news. Do they use the same sources for local events vs foreign events? Do they rely on blogs? What are the patterns? Let’s use for example the Gaza/Israel conflict, as it was covered enough locally and internationally. Do people cross check things through different websites? Or stick to the ones they know?
I’m wondering if Media Cloud will be able to analyze not only information in and around a news story, but statistics on the people who READ a story, pass it on, etc. It seems that in today’s world of immediate global communication, these figures are necessary to draw important conclusions about the relative influence of different news outlets. Do you have any ideas for obtaining or estimating the number of people who read about a given topic for each source covering it? Data such as unique visitors to a page and “send to a friend” emails is typically confidential.
Similarity of stories used to track wire services:
I recently found a computer crime related story in New Zealand which I wished to know more about the details. Searching the internet I found that almost all of the stories were identical with minor gramatical changes. Eventually I found that all of the reported facts were from a press release from the the FBI. I regarded this as a biased source. Almost all of the articles used facts from the FBI release, and most of them used identical phrases.
It made me wonder how much international news is simply grabbed off wire services, taken at face value and re-written to an editors preferences. I think tracking news as it spreads internationally, and comparing for similarity would be informative. It would indicate which papers verified/changed/added details and how often these papers did this. It could indicate the major channels of news dissemination around the world, it may allow users to more quickly find articles that actually added new/more information, and it may also allow a ranking of the quality of reporting from a specific source. In my previous case it would indicate that the information from a story set in New Zealand originated in America.
extensive keyword stores allow for broad applications:
On another note, I suspect that an extensive key-word store for each article would be useful. A lot of the above commenters wished to track word usage on a time-line; Africans in Europe/Financial crises/reporting on Israel. No doubt similar requests would arise in the future. A extensive store of keywords for past articles would allow future applications not yet envisaged.
I’m interested in the transition from traditional journalism to internet journalism, and how the internet user can “auto-detect” the professionalism and lack of bias of an online news source. One clue might be to detect and distinguish “words of opinion” from “words of fact.”
I’d like to understand how the semantic web will unfold, and what the first commercial opportunities are that it presents.
We are interested in determining how media coverage of education issues influences state education policy.
I like the idea of being able to see how the same global story is covered in various countries, societies. As we become a smaller planet, as we face common challenges which require a coordinated international effort, this sort of tool could provide an enormous boost to international understanding. Of course, this sort of international understanding, shared cultural knowledge and experience will make coordinated effort more effective.
Scientists are pointing to some pretty significant and urgent global issues- some have been around for many years and those of us more fortunate to live in more ‘blessed’ economic areas have been able to avoid the realities, the impact on our daily lives. This may change soon– Water is a primary example. And, as water becomes more scarce, global food supplies may also suffer… These are but two of many examples of the issues putting life on the entire planet at risk.
Anything this project can do to help promote international understanding- cultural exchange, sharing points of view on issues we have in common would be a wonderful help!
I’d be interested in knowing the differences in coverage between the Mainstream Media and the Alternative Media.
Il n’y a que l’anglais dans votre monde?
[At the moment, we're tracking primarily English news sources. However the system architecture generally supports any language and ultimately we hope to expand to many of them. - Ed.]
I would like to see a scientific based study on leanings of the news media from far left, left, central, right, and far right. Is Fox News really a far right Republican-oriented propaganda channel (with website bias too)?
I am interested in tracking source affiliations. Source affiliations are crucial for evaluating credibility, but no news story written in conventional narrative format can unravel the network of affiliations in the context of a narrative. Sometimes, if the story is a profile, the subjects affiliations will be extensively detailed, or pulled into a sidebar, but when the next story rolls around, and the profiled subject is now merely a named source. Source affiliations seem like an obvious resources for exploitation by semi-automatic and crowd-sourced projects.
I’d like to see all articles about the death of newspapers and the rebirth of news in other venues.
Love this concept..
I am interested and currently writing on how media messages reflect, refract or contradict political messages with a point of understanding how we linguistically understand policy, particularly education policy for me (words like accountability and such) but in all policy areas. Right now, it is exhausting to do, even using ATLAS.ti, because there is so much data just comparing a year’s worth of one politicians speeches and one newspapers coverage of a singular issue.
Having a “random” search function would be very interesting and fun — so that any three sources could come up for comparison. This could acquaint us with more sources that we were not already familiar with.
Simply great! Are there any plans for launching sister projects on non-English language media?
[Yes. Russian is likely, as well as other languages. -Ed.]
I appreciate the 24-hr time after the date underneath the name of the individual submitting a comment. For that time information to be meaningful (since your readers span the globe), it would be helpful to know what time zone the message came from. Or better still list the Coordinated Universal Time (Zulu time for pilots and military). This would require readers to know the positive or negative offset of their time zone in order to convert UTC to the reader’s local time.
In that same vein, I trust that as you collect news items from around the world, there is a time signature on each article. This would aid in tracking how news flows. For newspapers, it would be time hard copies first leave the press; online newspapers would have posting time, so too would blogs. Movies, television and radio programs, theater openings, books, speeches, magazine articles, press releases, journals, etc., would each have their own specification.
In this age of instant digital communication, time flies. Contrast that with a statement attributed to George Washington after Benjamin Franklin had been in Paris for about a year without any word from him: “Perhaps we should write him”.
Interesting, especially the insular world view of the blogosphere. Is the data available to the public. In terms of analysis how about thematic analysis? Semantic network analysis? Multivariate statistical modelling?
[Yes, we're just figuring out how to format it. Do you have thoughts? - Ed.]
Fun tool! I compared CNN with The Miami Herald. Looking forward to the day when we can compare by keyword.
[You can, at least if we understand you correctly. The "pivot" query will compare several sources based on the keyword you enter. -Ed.]
I’d love to track what the press is saying about our large public research university, and use that information to read between the lines of what that coverage might mean for our researchers, faculty, students and fundraising arm.
Frequency of use of keyword or phrase by source over a stated time span is primary interest. For example use by source of “gun” or “gun control” or “prevented attack” over the last two weeks. The results should be easily exported as a table.
Providing a data feed that can be queried opens up the possibility for the consumer to get involved with crafting the purpose and display. I would like to create a trend report that follows specific words (“Realtor”, “Real Estate”, “Foreclosure”, “Interest Rate”, “Bank Owned”, “FSBO”) from 3-5 sources over time. Are the use of these words going up? Down?
I would also like to compare markets. Possibly the same set of words but compare the Phoenix market to the Denver market so I would target publications in those 2 markets.
This is great. When do Canadian news sources come on board? I would like to be able to track how many stories are accompanied by video, audio, slide or data enhancements across media in various countries.
I have been researching the use of the morality frame in the media’s coverage of sexual issues, in particular sexual health. If possible, I would like to use the information here to expand my research and be able to see what words are associated with key terms, as well as how much media coverage on sexual health is out there in traditional media versus blogs.
@Djuana Stoakley
Djuana, you can definitely check out the American Time Use Survey. Good luck!
I have the sense that there is a lot more you can do to analyze text than just looking at the words. Words have meaning, but they are not meaning itself. Words are not subjects.
How deep can you drill with just words? What else can you do to analyze these texts?
[You could do an awful lot, and we're looking for folks with ideas. However, words are a useful starting point, and most of what you might do involves words or phrases. What do you think? -Ed.]
Do you hava an Ontology behind your data? Are you planning to join the Linked Data Cloud?