Difference between revisions of "Perfect 30 articles for CommonAnswers (Sept. 1-15)"
(→Reflections On The Task) |
(→Measurement) |
||
Line 144: | Line 144: | ||
* [http://c2.com/~ward/SearchRank/words.txt search words] for which we are now collecting search result placements. | * [http://c2.com/~ward/SearchRank/words.txt search words] for which we are now collecting search result placements. | ||
* [http://c2.com/~ward/SearchRank/order.cgi successful words] for which we now place in the top 100. | * [http://c2.com/~ward/SearchRank/order.cgi successful words] for which we now place in the top 100. | ||
− | * [http://c2.com/~ward/SearchRank/results results] as flat files for all queries. | + | * [http://c2.com/~ward/SearchRank/results intermediate results] as flat files for all queries (useful for double checking results.) |
Tip: Watch ''successful words'' to see if google has found the articles you have written and returned them as the result of a search. Click on a link to get a text file with historical results. Use Excel to plot column A (sample time) and column B (rank of highest placement) as a scattergram. | Tip: Watch ''successful words'' to see if google has found the articles you have written and returned them as the result of a search. Click on a link to get a text file with historical results. Use Excel to plot column A (sample time) and column B (rank of highest placement) as a scattergram. | ||
− | |||
− | |||
== Reflections On The Task == | == Reflections On The Task == |
Revision as of 14:43, 28 August 2007
DistributedPlanningGame / Create 60+ articles for CommonAnswers (August 16-31) (Your name here)
Contents
Why are we doing this?
- Begin creating content for this project to help generate traffic.
- Currently a lot of items being searched on AboutUs return very little or no information at all on these pages. We need to create at least 50 articles for the top searched terms.
Description
We've been teasing out ideas about how this project will proceed over the past weeks, but now we'd like to put a little elbow grease into the project and learn by doing. This first portion of the project is intended to be both a deeper exploration for new (and old) content hires (welcome to the team!) and a chance for all of us to collaborate on how these pages should look and where this project is going. We think that pushing through this first round of articles will teach us a lot about the process and allow us to refine it and the project even further. This is a great opportunity for all of the recently expanded content team to work together building our next resource and for our two offices to refine our collaboration process itself.
We've provided several examples of what we've come up with as initial drafts for some search terms. These are not perfect yet, but they do proviude a good idea of the framework in general. Portland will be refining these (and more) and fleshing them out over this work period. See the Portland Search Terms List below for updates on how our work is going and further examples in progress.
As we continue with this work ourselves, we'd like to invite 10 people from the Lahore content team to do 5 articles each over the next week. In writing these articles, please refer to the examples for a general idea of where we're coming from as far as format and style are concerned. That said, we can maximize the value of this project by making these pages what we each believe to be the best set of information for a given search term. We want these pages to collect and organize all web information on a term based on what YOU would find most useful for a particular search term. When we're done, we'll analyze on our collective work and reflections on the process and go from there.
The section below the Tasks list will provide a more detailed explanation of specific ideas for pages, but questions can be directed to ChrisBabson and ScottKeeler.
Tasks
Portland: Facilitate, continue work on example articles and more (5 each for ChrisBabson and Scott Keeler, and anyone else who is not otherwise over-occupied with content projects - add your name here...)
Lahore: Write 50 new articles (5 each)
Lahore Search Term List
- Category:Game Cheats
- Category:Obituaries
- Category:Cars
- Category:Pets
- Category:Song Lyrics
- Category:Baby Names
- Category:Family Guy
- Category:Flowers
- Category:Online Games
- Category:Simpsons
- Category:Anime
- Category:Name Meanings
- Category:Free Radio Stations
- Category:Banks
- Category:Lahore
- Category:Real Estate
- Category:Airline Tickets
- Category:News
- Category:Maps
- Category:Brad Pitt
- Category:Angelina Jolie
- Category:The Beatles
- Category:Pink Floyd
- Category:Lindsay Lohan
- Category:Avril Lavigne
- Category:Danica Mckella
- Category:Nicole Richie
- Category:Linkon Park
- Category:Justin Timberlake
- Category:Dance
- Category:Jazz
- Category:Environmentalism
- Category:Fashion
- Category:Magazines
- Category:Movies
- Category:Music
- Category:London England
- Category:New York
- Category:Los Angeles
- Category:Transportation
- Category:Health
- Category:Fitness
- Category:Soccer
- Category:Gardening Being worked upon... (would be good to redirect Gardening to this category also?)
- Category:The Matrix
- Category:Titanic
- Category:Casablanca
- Category:The Bourne Identity
- Category:Horticulture
- Category:Asia
Portland Search Term List
Newcomers to the project: talk to ScottKeeler or ChrisBabson about your article queue.
- Category:Portland Oregon
- Category:Paris Hilton
- Category:Weather
- Category:Athletic Shoes
- Category:Europe
- Category:India
- Category:Architecture
- Category:Goth Music
- Category:Painting
- Category:Allure
Some Other (substitutable) ideas - (ADD TO THIS)
- Category:Mysticism
- Category:Meditation
- Category:English
- Category:Spanish
- Category:Farci
- Category:Underage Drinking
- Category:Government
- Category:Headphones
Super Specific
If you're feeling experimental after your 5, try an iteration of any similar term to drill down to a subset. We'd like to see how this pans out for more specific category intersections.
Further Description and Discussion
The purpose of this weeks work is to attempt a slightly different approach to article creation. Instead of writing long articles with information from the web (often a redundant resource even if it is valuable), we're going to try to organize web search results into an article-esque form. The idea is to provide the user with human edited rankings/descriptions on a human edited article page. Our intention is not necessarily to be like these sites, but it might help to think of these articles as a combination of a Wikipedia and Mahalo page.
To start, get a sense for the top web results from Google and all other major search engines. The idea is to determine which of their results are actually the most relevant. The pages should include (by linking, with a short description) all the web information available, but should organize it based on your determinations of relevance. Make sure to keep in mind the broad sense of the term you're working on here so that any confusion can be avoided.
As far as tone, our idea so far has been to add some panache: we want these pages to have a clearly human feel (this is one of the values of a wiki after all). Refer to the Category:Portland Oregon page for a concrete example of this - the single sentence description up top provides the user with a sense of the flavor of Portland. This feel should be pervasive in the article and should apply to either single sentence or phrase descriptions for each link. Down the line, these descriptions may be pulled from the AboutUs page on that domain.
The article should have its own original content, but will not necessarily include a lot of text. The idea, again, is to organize and contextualize the information available on the web into an understandable format. The ultimate goal is to provide more comprehensive information. In doing so, we take advantage of all the domains we list by interlinking.
This brings us to some specific things we've noted so far. Feel free to add here if you like:
- 1) Treat each category link as its relative page name (i.e. Category:Game Cheats should include an article on the term 'Game Cheats' - redirect Game Cheats, GameCheats, etc. here). In this case, you would be doing an article on 'Game Cheats' that should organize all of the useful (and specifically not useful) information on this subject. Your results should be placed in order of relevance to the term - which pages do you think provide the best information on this subject? Be careful not to delve too deeply into information that could be contained in a potential subcategory (some of these search term categories are very large). The next step in this project will be to get more specific, so don't worry about getting too granular for now.
- 2) Organize the web results information into sensible sections. See the Mahalo articles on the same subjects for an idea of how things are organized on their site. Make sure that your pages don't imitate theirs, but add more content (describe each link by type of information or usefulness).
- 3) Links should be internal for now, and external only where they have to be (specific page or subdomain link instead of general site url) In the long run we will move toward a system where both are available.
- 4) Inculcate your article with some of the links from the category page you're writing on. We have a lot of information on long tail sites and this may be our advantage here. Your results should first pay attention to relevance, but also include these useful and hard to find long tail references.
- 5) Find and use any public domain pictures you can. CreativeCommons.org is a useful resource for this, as is Flickr (click on explore and then navigate to 'Creative Commons'). Make sure to cite your images properly upon upload in the comments section.
- 6) Link to any other category pages we have that might be or are subcategories for this term.
- 7) If there is any recent or past news that is relevant to your term be sure to include it.
Measurement
We've set up a script to measure the impact our writing has on the placement of AboutUs.org in response to Google searches for the above terms. As of Monday, August 27th, we fail to show up in the top 100 search results for any of the above terms. We will rerun the script every four hours and collect the results for each term in an Excel friendly tab-separated history file.
- search words for which we are now collecting search result placements.
- successful words for which we now place in the top 100.
- intermediate results as flat files for all queries (useful for double checking results.)
Tip: Watch successful words to see if google has found the articles you have written and returned them as the result of a search. Click on a link to get a text file with historical results. Use Excel to plot column A (sample time) and column B (rank of highest placement) as a scattergram.
Reflections On The Task
Add your reflections here:
I like the task, but feel that the categories we've chosen to experiment with are a bit too broad and we should select narrower ones which will be easier to write about since they are will have fewer articles and also be easier to measure from a standpoint of showing up in search engine results. - Ray | talk