The OpenAustralia/TWYF interface, design, layout and architecture present some significant usability and information consumability issues. This proposal seeks to identify those issues and provides a basic way forward for addressing those issues through what could be considered a fairly drastic overhaul. If such changes are not viable (and the outcome of the UI abstraction discussion will influence that) then there is scope for decomposing this proposal into small upgrades and modifications to improve usability at a more granular level.

This proposal is neccessarily critical of OpenAustralia in order to effectively highlight the issues, however bear in mind that I'm obviously a big fan of OpenAustralia so please no one take this personally or to mean that the entire site is a failure. That is not the case.

The issues identified through user feedback and expert analysis include:

* Unfamiliar terminology
* Lack of context
* Lack of organisation
* Lack of linkage between related information
* Pre-requisite knowledge of parliamentary workings

This means that to use OpenAustralia effectively a user must have knowledge of the mechanics of the Australian Parliament and familiar with the political and legislative environment at the time relevant to any hansard material being accessed and read. What is a bill reading and why are bills read several times? What is question time? What do all the different committees do? What is the background and context of everything?

OpenAustralia essentially consists of large numbers of isolated data with little metadata that make it difficult to present information in a meaningful way to users.

Take for example the first four Senate debates from Thursday:

# Notices
# Business
# Notices
# Business

What do these link to? Why are there duplicates? Then consider another debate further down the list:

HIGHER EDUCATION SUPPORT AMENDMENT (VET FEE-HELP AND PROVIDERS) BILL 2009; NATION-BUILDING FUNDS AMENDMENT BILL 2009; TAX LAWS AMENDMENT (MEDICARE LEVY AND MEDICARE LEVY SURCHARGE) BILL 2009; TAX LAWS AMENDMENT (2009 MEASURES; No. 2) Bill 2009; SOCIAL SECURITY LEGISLATION AMENDMENT (IMPROVED SUPPORT FOR CARERS) BILL 2009; SOCIAL SECURITY LEGISLATION AMENDMENT (IMPROVED SUPPORT FOR CARERS) BILL 2009; SOCIAL SECURITY LEGISLATION AMENDMENT (IMPROVED SUPPORT FOR CARERS) (CONSEQUENTIAL AND TRANSITIONAL) BILL 2009; Defence Legislation Amendment Bill (No. 1) 2009; TAX LAWS AMENDMENT (2009 MEASURES; No. 3) Bill 2009; FAMILY ASSISTANCE AND OTHER LEGISLATION AMENDMENT (2008 BUDGET AND OTHER MEASURES) BILL 2009; FAMILY ASSISTANCE AMENDMENT (FURTHER 2008 BUDGET MEASURES) BILL 2009; FAMILY ASSISTANCE LEGISLATION AMENDMENT (CHILD CARE) BILL 2009; PRIVATE HEALTH INSURANCE (NATIONAL JOINT REPLACEMENT REGISTER LEVY) BILL 2009; SOCIAL SECURITY AND OTHER LEGISLATION AMENDMENT (AUSTRALIAN APPRENTICES) BILL 2009

Wow. WTF? Click through and the actual information contained in that debate is far less than the ridiculously-long title.

Hansard is not designed to be consumed by the average citizen. It is designed to be a record of Parliament and thus is neccessarily structured as such.

For OpenAustralia to better meet the needs and specifically translate this structure to users' mental models we need to do some work.

I proposed a taxonomy and translation mechanism for clean URLs - see JIRA http://tickets.openaustralia.org/browse/OA-226

I believe we should take that idea and not just apply it to generating clean URLs but as one (of several) organisation methods for OpenAustralia so that users may browse debates by, say, "environment" rather than have to sift through other non-relevant debates like tax, migration, ozcar etc that they might not be interested in.

However obviously we can't expect all users will be coming to the site to look up a particular topic aligned to such a taxonomy.

We already cater for chronological and speaker - though I believe we can do this better, particularly putting debates within the context of time as topics of many debates span more than just the one day.

Sorry ... just got distracted with a thought of a "hot topics" page that shows trending topics in the Hansard, accompanied by hashtags.org style sparkline graphs. Not necessary - but cool nonetheless.

Anyway, so if we get that topics thing happening then we will be able to provide a very rough additional navigation mechanism between debates to link to "the next debate on environment" etc. Not sure how useful that would be, but something more specific like "This bill was also discussed 3 days later" and link to that.

We need to be generating abstracts for debates. Unfortunately that is a massive, manual job - it cannot be done programmatically. But it's something we seriously need to think about, maybe just identifying a couple of "hot" debates every day and adding abstracts. Will require a web admin interface to make this viable for the OA team. But how much more useful would the home page be if every debate had under it a short description of what the debate was about, the prominent speakers and their point of view?

Social network. Yes, OA is a social network - but only barely. We need to step that up, we need to increase the depth of user profiles and make interaction better in orders of magnitude. People should be able to comment on an entire debate rather than individual speeches. We need comment nesting, editing, community moderation etc.

The Wikipedia thing - it doesn't work for me. I was reading a debate the other day that had two Wikipedia links in it, both applied incorrectly. We need to have our own OA wiki that the community can create articles on and then the automagic linker can link to those when it finds a match.

Guidance. We need to provide contextual information on the procedures of Parliament in-situ where people are interacting with the site. So if someone's looking at a question time hansard then information on what question is should be easily accessible. We need to accommodate users who have absolutely no idea how parliament works - and I know when I first accessed TWFY when I joined the OA project 18 months ago it was very daunting. All these references and words that made no sense to me. What do they mean? What is their relationship with each other? How do I get an idea of the big picture? We need to make sure we can provide that big picture and context to users through the IA, navigation, contextual help, explanatory guides, diagrams etc.

Also with the incoming transcribed RMI data there will be more opportunities for cross-referencing data ... I'll leave that up to the devs as I'm not aware of any particular such information need but basically as many such opportunities that can be taken advantage of there without negatively affecting the user experience will help provide interest and context. Who said what and when, when was what said, who is involved with x, when was x discussed etc. There's going to be some real gems in there re transparency through data aggregation ... and that's the information that's going to equip citizens with the information they need to be fully informed about what MPs are doing ... without having to try and link all those points in their head.

I want to see the home page changed. I don't want to see a list of the titles of debates from yesterday - though obviously we need to retain that navigation scheme for people who want to just want to scan and access stuff discussed yesterday. And any other day. I want to see it broken down into topics, into abstracts, prioritised by the community so that the most interesting stuff is brought to the top and less interesting stuff is pushing below the "home page exposure" threshold onto a "more ..." page. I want to see the social networking stuff made more prominent instead of just the last ten people who commented. What is really happening in the backchannel? What's the underlying sentiment? Which MP is attracting the most criticism? All these questions are really hard to get answers to without reading everything on the site ... and it doesn't need to be that way.

Anyway - that's me for Part I ... share your thoughts! If you think it's worthwhile then a few of us who are willing and available should get together for a brainstorm, nut this out and then start prototyping some new designs. I expect that implementing this fully would require an overhaul of almost every page on the site.