Multiplicity is a platform to simplify news consumption by giving users the power over what, when, where and how they consume content. It aggregates news into a common format so users can remix and visualise the information they want, to uncover what's really happening.
How it works
Multiplicity converts news content from multiple sources
Multiplicity aims to address these problems:
1. news is published in many different data formats, making comparisons difficult.
2. there's a vast amount of news, from both official and social sources, impeding comparison and comprehension.
3. the optimal news experience differs across platforms and users.
Multiplicity addresses these problems by:
1. converting data from specified sources into a common format
2. enabling users to drive the process using search terms or topics, then:
(a) use the common format to generate comparative or aggregative visualisations, such as word clouds, bubble charts, sparklines and more; and
(b) filter their results on all available variables in the common format, from sources to dates to location to content type.
3. leveraging context-aware knowledge to guide the user configuring their experience for their current platform.
At each point in visualisation, users can choose whether to aggregate or compare sources, so they can trace patterns emerging from difference sources or see the overall picture. Naturally, not all sources will have a complete metadata set, so it will be transparent to the user how many items from each source are being included in a given visualisation.
The Multiplicity platform is modular, with users able to save, export and import different newsviews, contribute new visualisation modules, and help drive the data conversion process. Exported/shared newsviews are embeddable in other sites, enabling users to showcase news curation by themselves, friends or favourite journalists.
The community will be invited to contribute to the data conversion, by using a simple graphical tool to map metadata from a given news source to Multiplicity's common data format.
This enables Multiplicity to overcome the semantic gaps inherent in aggregating news items.
It's essentially many eyes + google news + social media aggregator with added context awareness.
1. Implement and test the format washer to convert news formats.
2. Expand prototype to encompass all available filters and visualisations, and iteratively test on newsroom journalists.
3. Recruit external users (developers and end-users) for early feedback alpha programme.
Integrating this into a newsroom (and work devices outside the newsroom) is straightforward: it runs in a browser (on multiple devices) and supports a range of news curation activity, either for research or as an end in itself.
Journalists will be trained and encouraged to share their newsviews. Ideally, I would spend time with several key journalists to analyse how the tool can support their job, then generate and export matching newsviews for each person to underline its usefulness.
The shareable nature of newsviews means everyone can collaborate in investigating news patterns.
Sharing is key:
- Really need visualisation / interface examples.
- Can the metadata conversion process yield a useful working set for visualisation?
- Can a browser-only solution extract data from a variety of sources fast enough to give a responsive interface?
- Should user preferences be stored on devices or in the cloud?
- Big UI challenge to make complex visualisation tools work well on small devices, regardless of context-aware support.
- Seeding sufficient sources to start a critical mass of interest
- legal issues
- Right balance between exposing variables and making the interface usable
Plans to resolve issues
- Establish appropriate common format to capture the maximum news metadata
- Create simple developer tool to support the conversion of source metadata into the common format.
- Furious prototyping and iterative testing.
- Bat eyelids at excellent front-end developers.
- Consult lawywers.
Multiplicity could also be extended by adding more subjective metadata measures and collating user ratings of news items, for example on left- or right-wing bias, or on trustworthiness.
Why newsrooms need Multiplicity
News delivery has traditionally been a top-down affair: Multiplicity gives users the tools to configure their news for their needs, which vary not only from user to user, but also for an individual user from hour to hour, depending on their current context, e.g. what device they are using; the time of day; where they are. It's a bottom-up platform that enables end-users and journalists to obtain the news information they want, in the format(s) they want, when they want.
Humans are good at recognising patterns, but the sheer wealth of news generated each day is overwhelming without aggregation tools. Multiplicity aims to support journalists to explore emerging patterns from a variety of news sources, and give them the power to remix the information.
Ultimately, a Multiplicity newsview, an item of curated news, is news content, and has value in itself.
How Multiplicity facilitates telling a story / who will benefit
The ultimate aim is to enable users to discover news patterns. Journalists can use Multiplicity to: track emerging news; investigate a particular topic; investigate sources to check authenticity; curate and share their own views of world events.
Journalists and end-users alike will benefit from having a powerful visualisation tool to make sense of a plethora of news content. Organisations can use it to improve the transparency of the news generation process. Finally, user data from context-aware modules can feed back to advertisers, giving fine-grained information about user activity.
Leveraging off existing apps
- The modular nature of Multiplicity means that it launch using visualisation packages like d3.js and infovis, and that it is extensible to incorporate new packages.
- The FOAF functionality leverages information from users' existing social networks.
- We would investigate using something like Mashape to handle API collation and Scraperwiki to generate APIs from Plain Old Pages.