This is an archive of my original “Next-generation political blogging” Writeboard, which a few of us worked on between 31 May and 20 July 2009, and which set the early direction for Poblish. Hopefully it gives you a good idea of why we did things in the ways we did, and why we still reckon we were on to something!
Preamble: types of political blogging:
- Political gossip; personality-related. Not relevant to this discussion.
- Reportage, scoops. Not relevant to this discussion.
- Pure promotion of a party or organisation. Not relevant to this discussion.
- Game-changers; bloggers who seek to make a decisive intervention at a crucial moment. These may not be just conventional blogs but they can be activist groups that have grown out of, or are based upon, the idiom of blogging
- Issue-related political blogging: from philosophy down to policy tweaks. Aimed at policy-makers.
- Outcome-related political blogging: social activism and campaigning. Aimed at individuals and groups.
Current problems
- The yawning gap between posting and policy-making.
- General lack of connectedness, and especially interaction between bloggers: efforts are duplicated, arguments are missed.
- The need to encourage greater responsibility among bloggers and readers, closing the gap between the two. Conversations and collaboration, rather than assertions and ‘feedback’ via the (anonymous?) comment box.
- Resolving issues is difficult: how can we encourage people to accept alternate views, or admit they have been wrong?
- Consequent lack of respect for bloggers among policy-makers, and alienation among bloggers.
- The crudity of mainstream blogging platforms, and the lack of specific tools for political bloggers.
- The difficulty in effectively mining political blogs (in bulk) for relevant information – Google is hopeless here.
- The need to be able to hold all involved with political blogging (politicians, bloggers, readers) to account for their views, and to make it easier to track their views over time.
- For each politician (or ‘political actor’), to increase scrutiny by leveraging all publicly accessible sources, not just their own blog.
Proposal
- More, and smarter blog aggregation.
- Less monolithic posts: tagging, categorising, and commenting on sentences/paragraphs. Posts and sections can be marked as ‘duplicates’, purely personal, shills, rhetorical, etc. All tags are marked with a username and timestamp.
- Take advantage of manual and automatic (we must have an excellent algorithm) tagging/categorisation.
- International and multi-party.
- Goal-driven blogging: focussed on ‘ideas’ (multi-jurisdictional) and ‘policies’ (within jurisdiction).
- A ‘resolution mechanism’.
- Greater responsibility: no anonymity; ‘political actor’ concept includes bloggers, commentators, and readers.
- All Users have public profiles (including affiliations, interests), though contact details are hidden.
- Groups can be formed on the basis of profile settings, though partisan groups should be avoided.
- Interoperability: we import existing blog feeds and all publicly-available sources (e.g. speech transcripts from Hansard, TheyWorkForYou, etc.)
- We maintain a time-line for each User.
- Greater responsibility II: ‘force’ users (including politicians) to confront their contradictions, and explain their changes.
- We export feeds for each user, idea, jurisdiction, policy, tag / category.
- Realistically, this requires a new, purpose-built site.
- We encourage Users to turn off commenting at their own blog.
Use Cases
- How will political bloggers actually use / switch to this technique/platform?
- Have to accept that some political bloggers just want to write, copy, or campaign – rather than think or solve.
- Process of blogging should not be dramatically different, or too much more difficult, if we’re to persuade bloggers to use us. Needless to say, we should be giving them cool stuff / insights / rewards they can’t get from existing blogging.
Limitations & Concerns
- We can’t expect to solve all the problems of political engagement here.
- Competition. Not in the UK, surely…?
- Commitment, support, funding.
Components / deliverables
- A list of ‘jurisdictions’.
- A database with Users, who have Profiles, and (optionally) Feeds (political blog RSS feeds, TheyWorkForYou Hansard transcriptions, etc.).
- An aggregator, that feeds off that database.
- A component that categorises and perhaps indexes each post – quite possibly at a paragraph level – then storing the results.
- A way of visualising the results: multiple posts/sub-posts that fulfil the user’s criteria, displayed simultaneously.
- A web-site – or WordPress plugin – that provides a time-line for each User.
- A log-in facility for all Users.
- A section of the site – or WordPress plugin – that allows Debates to be created, managed, persisted, with reports, notifications, etc.
- A way of randomly selecting Users to become Jurors.
- A way of crediting Users for participating, and penalising them for failing to meet their obligations.
- DebateGraph integration?
- Outputs: aggregated feeds
- Outputs: policy debate resolutions
- Outputs: a way of ‘mining’ our ideas / data.
- Outputs: reputation-building, public profiles, User badges, etc.
- Finally, a way for a Use to connect their own blog to one of our Debates, rather than use conventional blog commenting.
Potential collaboration partners
- Individual bloggers – advice.
- MySociety – consultancy, project management.
- ??? (UK, Europe, USA) – funding, advice.
Next Steps
- Build interest and get ‘buy-in’.
- Create a Ning group.
- Implementation!
- A catchy name…
What do we have? @ 20 July 2009
- A search-engine for political blogs.
- Users are political ‘Actors’, each with their own customisable, searchable Profile.
- An opt-out system: the list of Actors is maintained by us.
- Users log in using their Google (including Yahoo, OpenID) or Facebook accounts, and claim an available Actor (validated by emailed code).
- Actors’ blogs are ‘scraped’, and feeds subscribed-to as soon as they are claimed.
- Extremely fast (indexed) searches, filtered by content, tags, categories, and Actor.
- Match snippets are shown, with links back to the source article.
- Visualisation of search matches on a timeline (see here), highlighted according to relevance (based on content and tag/category matches).
- A hierarchical ‘tree’ of terms is generated, allowing us to determine if a particular term/word/phrase is political, economic, etc.
Technologies used: Java, Hibernate, Lucene, Compass, XPath, AJAX, Prototype + Scriptaculous.
Details
Users
- Opt-out, not opt-in. We’ll pre-populate the site with as large a number of political feeds as we can (manage), and create ‘Actor’ entities for each (individuals and organisations). These will be claimed by the individual concerned over time. Some Actors might only ever be place-holders, and might not necessarily have their own feed.
- Users log in/connect with LinkedIn, Facebook, etc. APIs.
Resolution mechanisms
Q. How are disputes ever resolved?
A. By presenting contrary evidence and relying upon people’s intelligence/openness/honesty to adjust their own position. There can’t be any compulsion. Can be face-to-face, can be online too.
Resolution: ‘Debate’ concept
- A ‘space’, with a page. Users can participate, or subscribe to view updates. Can belong to a single jurisdiction, though would ideally be open to all.
- Each user who enters a debate declares – in public – a statement, and rejection criteria. This will appear in their profile. Administrators will reject (and penalise) entries that are unfalsifiable.
- Users are credited for changing their mind. Subscribers are notified.
Resolution: ‘Juries’
- We randomly select ‘jurors’ from among all users for the jurisdiction(s) in question. Aim is to resolve disputes, to throw out bad arguments, but not be directly politically involved. This would be fun, and tick the ‘responsibility’ box.
- Jurors would be credited for getting involved.
- We could require 12-0 or 11-1 votes within a Debate in order to empower independent/disinterested thinkers.
- Participants would have to comment on the results of the debate (rather like eBay): and would be penalised for not doing so, or (perhaps) for not accepting its results.
Other thoughts…
DebateGraph
- How can we interact with this?
- There doesn’t seem to be an API (yet?), but in an ideal world we could automatically generate DebateGraphs, and allow our users to log in/use that site seamlessly.
Implementation details
Potentially any or all of:
- A standalone website – a viewer, ideally. We wouldn’t want to have to create a fully-featured blog editor, which would put us on a collision course with WordPress et al.
- A suite of WordPress plugins.
Aside: Rating politicians
- http://r8yourpolitician.co.uk/ shows us exactly what not to do (in many ways).
- On what (fair) basis can this be done?
- Quality of casework?
- Appropriate general level of communication?
- Timely response to individual questions.
- “Keeping promises”? Unfair. (a) nobody makes promises, and (b) changes are legitimate, but should be explained (see above).
- We can hold politicians to account using the categories and timeline. Our system should be the best ever way of comparing positions over time.
Miscellaneous
This should also accept that:
- Personal, private views.
- Personal, public views.
- Party/Governmnent public views.
… intersect, but are not identical.
Also distinguish between:
- What people state that they believe, on a particular idea/policy – which might very appear definite, but which will fluctuate.
- What people are prepared to accept, given the room for manoeuvre they give themselves on each idea/policy, related to their level of doubt, priority, commitment, etc. These boundaries will change much less often.
Further thoughts (7-11 June, following on from Miljenko’s comment)
‘Ideas aggregator’ and ‘Academy of Progressive Thought’
- Are ideas much use without a ‘scientific’ scruitiny: context, evidence, experiment, etc.?
- Leads back to DebateGraph, I think…
Community Activism done better: existing approaches
- Is Facebook’s ‘Causes’ helpful? Debate and supporters brought together, with an incentive to attract others.
Distribution models…
- Competition for eye-time. What can we do in a world of established blogs, and aggregated Facebook and Twitter feeds?
- YouMagz…