Environmental Scan
This page is for visual ideas from other projects for the new TAPoR. -- GeoffreyRockwell - 28 Oct 2010 *Amazon’s site could be an ideal model for a redesign (and Professor Rockwell did also express some interests in Amazon) but Amazon also recently went through a redesign as well mostly on their landing page. Instead of a cluster of categories in a grid layout we now have larger pictures of items representing their individual stores. However, clicking on their “books” category directs one to their old familiar layout. (essential) *The particular thing about Wolfram Alpha that might be relevant to the Redesign is the way it organizes its search results. For instance I searched for the word “tools” which was able to provide the different definitions according to its different senses (mainly using cues from Part of Speech), then it provides an etymological background in another boxes; the inflected forms of the word in another box; typical frequency of the usage of the word in another box; it even proceeds to provide synonyms of the word tool and other trivia information (like rhymes, anagrams, scrabble score and crossword puzzles) not forgetting the broader semantic terms for the word tool in neat separate box categories. The results page of Wolfram Alpha may be useful to the Tapor redesign. (essential) *I searched for the same item “tools” on the Duckduckgo site and it proceeds to hand over any fine-tuning to me by stating boldly: “Tools can mean many things. Click what you mean by tool.”—with options for both web and dictionary links. Duckduckgo mashes a lot of its results from other search sites, Wikipedia and dictionaries to help disambiguate the meaning of a search. If Tapor tools decide to integrate data from other sites, I felt it was useful showing a sample of how one project does it. Even Wolfram Alfa possibly integrates data from Wordnet or other types of dictionaries when trying to provide a word’s synonyms for instance. (essential) *Amazon is mentioned again because it has a very useful recommendation engine that shows what other users are already searching for or buying in an attempt to persuade the surfer to also tag along and buy its most popular items. The recommendation engine proposed for Tapor would glean both UI insights at the frontend and the possible types of log analysis and cookie/IP tracking that makes this type of feature possible. (essential) *Linguistlist was referenced here primarily because it has been able to create an online community based on its projects with partners and provides a gateway to those partner projects. I wanted to draw your attention to the link that says “Ask a question” which instantiates a drop down menu that directs me to be more specific about whom I want to address the question to. (essential) *The Natural Language Toolkit is mentioned as a good open source model to tool development that could be useful for Tapor. With NLTK anyone can view the source code of the current NLTK tools and decide to contribute to a project’s development on Github for instance. If the originally authors of the NLTK like the improvement or new tool, they can decide to merge it in future versions of the NLTK. I am not so sure if the Taporware tools can be open sourced and if licenses would allow people use and modify code. But it has helped a lot of other tool projects so far. (optional) *Reddit and Digg were mentioned as two different approaches that have been taken to engage users towards news items. They have different voting systems: The Reddit approach is a lot more transparent because it allows users to up or down vote items using opposite facing arrows. Digg only allows upvotes and its voting system is a bit opaque. When extended to the Redesign, if users will be allowed to vote on popular tools for instance in a Reddit-like transparent manner, how does it benefit the larger community of tools users? Will developers of those tools feel slighted? Will a more polite Digg approach be more appropriate in that context? (optional) *The Youtube page was mentioned to illustrate a different approach to implementing the “vote” icons in the form of a thumbs up or thumbs down “like” icon. After reading so many Youtube comments it is interesting to note how these icons generate so much emotion and heated debates especially when fans of a particular program/music/person get aggravated because other people disliked the item they passionately liked. Text analysis tools may not generate the same type of heated debates but the Youtube like/unlike icon was mentioned as a reference to how it differs from the up/down arrow approach. The neat grid layout of Flickr was worth a mention here because of the way it uses tag clouds to both hint popular items and use them as same as category filters. (optional) *Postup is mentioned as another way to bring the various Tapor tools to the view of the user by having it simulate a scrolling feed of tool icons and descriptive text in a box. This could be used as way of bringing together the various tools to the user’s focus in a tight unit (the box), and possibly enhancing user engagement. (essential) Pageflakes
http://www.pageflakes.com/community/content/flakes.aspx
Nines Search Results
http://www.nines.org
http://dev.sencha.com/deploy/dev/examples/
Wolfram Alpha
Wolfram as a “computational knowledge engine”, it has also been regarded as a possible competitor to Google. It uses ontology and human curated data to return an answer with graphs and other representations.
http://www.wolframalpha.com/
Wolfram Alpha Results Page
Its ontology system also affects the way results are displayed in category boxes.
duckduckgo Results Page
Another engine that attempts categorizing according to subject area is Duckduckgo. Its results are a mashup from many search engine sources and Wikipedia.
http://duckduckgo.com/
Amazon
Amazon also tries to categorize items on its page.
http://www.amazon.com/
Google Books
Google also categorizes its books according to subject area and other criteria, including Trending Topics. It lists book items in rows of icon tiles.
http://books.google.com/
Linguist List
Linguistlist hosts an array of projects. Notice that featured projects get displayed in a bold banner at the top center of the screen to get people’s attention. The links below the Technology heading repeat the contents of the banner for emphasis. Links lead to research projects and tools developed by partners.
http://linguistlist.org/
NLTK
The NLTK site is a great resource for Python related Natural Language Processing Tools and Projects.
http://www.nltk.org/
NLTK Project Ideas
The NLTK ideas page tracks the development of various projects.
Notice Reddit's simple voting system.
http://www.reddit.com/
digg
Notice how Digg unlike Reddit has no "down" voting arrow.
http://digg.com/
Youtube
The “like” icon on Youtube as in everywhere, is a kind of vote where people indicate what interests them.
http://www.youtube.com/
flickr
Flickr collects photo image items and organizes them in rows of tiles. A Flickr tag also functions as a filter when clicked on. Notice how Flickr enables people to share photo sets with a simple widget.
http://www.flickr.com/
postup
Postup makes badges for blogs and in this particular case depending on the subject matter of that blog, it simulates a scrolling feed of icons and texts in that box.
http://techcrunch.com/tag/android/
