Proposal

High level description of our idea

Our idea creates a space for people at various levels of clothing/fashion sense, knowledge, and interest. It takes fashion in a practical sense more than a “glamour/model” sense. Our network will center around the ability to “bookmark” clothes by certain tags and categories.

  1. Find an item on the web or from someone else’s profile
  2. Add it to your profile
  3. The item (a pointer to the URL and a picture) will be saved to your profile
  4. Track your items, view others’ opinions, combine it with other items to make an outfit.

People will be able to leverage the social network to learn new styles, but can also be more practical: what are the best things to wear in various situations?

  • Work/Interviews (and doesn’t this depend on the profession?)
  • Other events or occasions / Everyday street
  • Mood
  • What are the best things in different weather?
  • Location: trends, available clothes, etc.

Users will be able to make smarter clothing and fashion decisions, and they will also have an easier time acquiring new things.

.

Our Target User

Our target user would be people who have needs to keep up with the current trend, who have trouble dressing themselves up and need advise from more experienced people, and who want to keep track of items looked online.

.

Example Usage Scenario


Mandy just got an internship as a user experience designer in Detroit and she is really worried about what she should wear to work at her first day. She emailed her mentor, who replied, “Oh, don’t worry! Business casual would be fine.” She was confused by the term “business casual” and started searching online. However, she was not satisfied with the results returned by Google image search because the look really ranged a lot. She continued her search and found an interesting website that asked her, “What do you want to be today?” or “What type of work do you feel like doing today?” She typed in “user experience designer” out of curiosity and got 26 returned results in 2 categories. Ten were single items and the rest was a combination of top, outerwear and pants. She then checked the faceted search items: “business casual” and “Detroit”, and narrowed the results to 17. She browsed through some of the results and got a sense of what business casual would be like in her industry and work location. She found some of the content were actually generated by users from Detroit, so it not only catered to the look of business casual, but also was suitable and comfortable to wear for current weather in Detroit. She felt this is a good tool for people to dress more properly to fit in certain circumstances.

Being happy with the result, she then created her account and filled out the profile form. After she logged into the system, she immediately got some recommendations based on her preferences and profile information, such as her weight and height, favorite brand, acceptable price range, and preferred style. She added Avril Lavigne as her “fashion neighbor” because she likes her a lot and would love to learn how to dress the way she does.

Later in the day, she happened to hear her friends who were using this system saying that there is an easier way to get better recommendations. With a delicious-like browser plugin, she can browse and search as she usually would and bookmark images she likes along the way. She felt it is not only easier to get better recommendations, but also simpler to keep track of items looked online.

.

Critical features

The fashion system will require a number of various features, most importantly, the ones below:

  • Indexing, “bookmarking” system. This is a way to store clothing items in a user’s virtual closet. They can choose to get items from the site itself or from anywhere else on the web.
  • Feedback/ranking/hype mechanism. This will provide a way to signal which products and items are the most popular on the site, what people are interested in buying. People can also use it to give feedback on peoples’ combinations & outfits.
  • Tagging & categorization. Fashion is very context dependent. People will be able to see choices based on certain interests and criteria. Does someone want to learn more about work clothes, but in a particular industry, and even a particular location? What is “business casual”?
  • Recommendations. Based on the categorization and hype mechanism, users can receive system recommendations that can bring new items to the user’s attention.
  • The “draw” or intrinsic value. The bookmarking can simply help people remember where they found certain clothes. People might just want a little fashion advice. Others will want to show off.

.

Platform choice

Website + possible Firefox or Google Chrome plugin

.

Competitors

1) Svpply.com

  • not just clothes
  • categorized by gender, price range and product type
  • ability to follow users
  • invitation only
  • item pages include, “other items you might like”, other users who like that item
  • not clear if it has more personalized recommendations if you are logged in or not
  • not clear how images/item info is being captured or entered – there is always a reference link back to the blog or store it came from

2) ShopStyle.com

  • browse items of clothing by category (pulled from retailer’s websites)
  • users can combine items to put together “looks” saved in their “stylebook”
  • ability to follow users
  • 3 heart rating (like it, crave it, love it), tags, comments
  • groups that act kind of like flickr groups – just aggregate members looks
  • does not appear to have personalized recommendations
  • “related looks” and “see more” suggestions appear to be based solely on categories
  • no demographic data gathered in profile

3) Other sites of interest:

.

Timeline

January 21 – Proposal Presentation [Due Date]
January 22 – Proposal [Due Date]
January 27 – Specifications
February 3 – Database Schema, Basic DB Implementation, User Profiles Capability
February 11 – Five Informal User Interviews
February 18 – Lo-Fi [Due Date] – Mike
February 18 – Complete Database Implementation
February 25 – Manual Contribution Functionality
February 25 – Test Plug-In
March 11 – Draft Design of Recommender System
March 11 – Architecture [Due Date] – Kate
March 12 – Architecture Report [Due Date]
March 17 – Final Plug-In
March 24 – 400 Records Mined
March 25 – Business/Sustainability Presentation [Due Date] – Kerry
March 26 – Business/Sustainability Report [Due Date]
March 31 – Implementation of Recommender System
April 7 – User Interaction Implementation
April 20 – Testing Benchmark
TBD – Final Report – Natalia