December 27, 2016
It’s been almost a year since my last blog post, so I thought it might be time to provide an update for anyone interested in what’s been happening with the Structured Stories project.
During the first half of 2016 I focused on research work, much of it associated with my fellowship at the Reynolds Journalism Institute at the University of Missouri. Early in the year I used the WordSmith tool from Automated Insights Inc. to demonstrate use of the Structured Story database to drive natural language generation – so-called ‘automated journalism’. Later in the spring I worked with a University of Missouri team to conduct a broad survey of user comprehension of structured stories, using a greatly simplified user interface and a small set of stories tied event-by-event to text articles. Publication of academic work has continued past the end of the fellowship, including delivery of a joint paper at AEJMC in Minneapolis in August, and delivery of an overview paper at the Computing News Storylines workshop in Austin in November, titled “Computable News Ecosystems: Roles for Humans and Machines”. Other papers about aspects of the project, by myself and others, are in progress and by the time the dust settles there should be another 3 or 4 published papers about the project, touching on reporting efficiency, story consumption and automated journalism. More on that to come.
But despite this academic success, and despite the substantial interest in Structured Stories from the structured journalism community, it has become clear to me that the journalism world is not well-suited to long-term research and development (R&D) projects. Journalism has no tradition of reimagining its fundamental components and assumptions, no institutions charged with exploring technology-heavy alternatives and, critically, no appetite for investment in work that might pay off years in the future. Fortunately, however, there is another domain that shares most of the characteristics and challenges of journalism but which does have a long tradition of supporting long-term R&D – the intelligence domain.
The intelligence community’s essential task is to gather information from a wide variety of sources and to refine and contextualize that information into products that help decision-makers understand a complex world. Despite recent controversies about access to communications metadata, the overwhelming majority of information processed by intelligence agencies these days is actually open source intelligence – ‘OSINT’ in the acronym-heavy government vernacular. The intelligence community employs tens of thousands of analysts to find and organize all this information, and relies largely on text documents to synthesise and communicate its work. Its workflows and processes parallel journalism in many ways, and the too-much-text problem is becoming as critical in intelligence circles as it is in news. Unlike the journalism ecosystem, however, the intelligence world has the ability and willingness to fund long-term R&D of technology-based solutions.
Beginning in mid-2016 I have therefore been adapting Structured Stories for use in intelligence applications, funded by a small research grant and aided by a research company with experience in government R&D funding. Although it is still early in this transition, results have been promising and the prospects for further funding for 2017 and beyond appear encouraging. Much of the functionality addressed by this new development work is directly applicable to journalism, and it is my intention to incorporate it into a journalism-facing product sometime in the future. I don’t expect to publish anything about the intelligence applications of Structured Stories until at least late 2017, but the ‘Computable News Ecosystems’ paper mentioned earlier describes the general system pretty well.
Developing an alternative to text articles as units of news is an audacious goal, and will likely take years, and a small but necessary detour along the way may be the quickest way to reach it. I am still deeply committed to offering a data-centric alternative to our rapidly collapsing text-centric news ecosystem. I intend to keep the existing Structured Stories site alive, and hopefully maintained, and I intend to remain active in structured journalism forums and conversations. I invite anyone interested in the project, or interested in helping out, to get in touch at any time.
January 5, 2016
The Structured Stories project wasn’t intended to last as long as it has. It’s purpose has been to explore whether or not it might be possible to apply computational narrative techniques to general news stories and, if so, to demonstrate that approach publicly. When I began the project, two and a half years ago, I had no idea what would be involved. I had a clear understanding of the devilishly difficult situation facing ‘post-industrial journalism‘, a frustration with the lack of proposed solutions, a background in data modeling and a keen interest in analytical approaches to ‘story’. My goal was to propose and demonstrate an original and plausible alternative to the doomed, article-centric approach to news.
It’s been a long and often challenging road since then. It took almost a year of iteration through data models using a research prototype before a stable representation of structured events and narratives emerged. It took almost another year to build an operational prototype capable of being used by others for reporting into structure and enabling anyone to explore structured narratives. And it’s taken more than six months of reporting, in three separate experiments with a total of 10 reporters, to start to understand the reality of reporting everyday news stories as structured events and structured narratives.
I’ve learned an immense amount along the way, mostly from the many talented and knowledgable people I’ve had the good fortune to meet and work with. I’ve learned how to think about structured journalism and structured narratives, where my approach fits within these fields and why there are particular opportunities from radically changing the paradigm of news. I’ve learned how to talk about the approach, how to work with journalists, and how to navigate the cultural gaps between journalism and technology. Above all I’ve learned that it really is possible for journalists to capture general, ordinary, everyday news as structured events and narratives. The constant theme along this path has been simplification, and I have a growing sense that the approach seems initially complex only because of the need for an unfamiliar perspective.
So what’s left to do? One major task remains – to create a taxonomy of event frames within a significant domain of news. The Structured Stories event frame library already contains almost 600 event frames – abstractions of ‘types’ of news events – and it has become clear that organizing these frames and standardizing their use is critical for applying this approach in real workflows. Building and publishing a taxonomy of event frames for news in a domain – specifically the local and state government domain – will complete the end-to-end demonstration of structured narrative as a viable approach to reporting and publishing news. There are many other tasks remaining. Loose ends abound, as anyone using the Structured Stories platform can see, but fixing those merely requires the time with which to attend to them. Understanding the event frame taxonomy is the last remaining area of exploration necessary before I can confidently propose a design for a real-world product.
October 6, 2015
Focus and engagement are both powerful things. Sometimes it’s necessary to be heads-down on a hard task for a sustained period. Sometimes it’s necessary to engage broadly with others; to communicate, to demonstrate and to collaborate. For Structured Stories the last three months have been about engagement.
The reporting phase of the Structured Stories NYC project completed at the end of July, and the reporting team succeeded in capturing more than 60 stories on local government topics in New York City. As the first independent reporting effort in a structured narrative format, the project produced a lot of findings, which will take time to fully understand. A paper on preliminary results was accepted for the ‘Computation + Journalism 2015’ symposium at the Columbia School of Journalism in New York. The key result is that the approach worked: Ordinary, everyday news stories were reported and published entirely as structured events and narratives by non-specialist journalists. I believe this to be a very big deal.
The next phase of evaluation is now underway, and includes a second, more narrowly focused reporting experiment covering Missouri state government and supported by the Reynolds Journalism Institute at the University of Missouri. The goal of this experiment is to understand how to use Structured Stories to cover long-running state and local government stories, and to build a library of event frames to do so. A separate experiment testing the consumption of news from Structured Stories is planned for late 2015 or early 2016. More on these experiments soon.
Other activities in the third quarter include a visit to NCSU in North Carolina for an invited talk in late July, an ongoing experiment on structuring pursuit stories with a major Los Angeles newsroom, work on a more consumer-friendly user interface with the Reporter’s Lab at Duke. The quarter also saw various articles and podcasts about Structured Stories, an unconference panel at ONA15, several posts on the RJI website, and lots of interaction with a diverse range of people on an even more diverse range of topics. It has been busy, fascinating and fruitful, and I’ve learned much.
But there’s lots to do, and increasingly it will require focus. The Structured Stories platform, which has served well for initial reporting experiments, must be upgraded based on those early findings. Many bugs need fixing, many elements of the API and UI need simplifying, features need to be completed or improved, and workflows need to be streamlined. Also, demonstrating the feasibility of Structured Stories is not the same as demonstrating its usefulness. Finding an initial niche for the technology and exploring various operational and business models will be priorities towards the end of 2015.
Much has been done, but it’s still very early days for Structured Stories and there’s much to do. If you’re interested in getting involved then get in touch.
July 7, 2015
It has been an intense time for the Structured Stories project – mostly due to the ongoing Structured Stories NYC experiment. And while I haven’t been very active on this blog there has been lots written about our experiment on the RJI blog, in a Nieman Labs article, and on the Reporter’s Lab blog, which is becoming a really interesting record of the experiment.
A deluge of excellent reporting has been pouring in from the impressive and productive team in New York (Ishan Thakore, Natalie Ritchie and Rachel Chason), along with a deluge of revelations, clarifications, consternations and realizations – all valuable learning that could only have resulted from real reporting. The quantity of rich information being generated about the reality of structuring news events and narratives far exceeds what was available before this experiment, and will take months to fully digest. This is the first time that Structured Stories has been used in a production setting and, while many bugs and inadequacies have been revealed, the team has nonetheless been able to successfully use the software to continually capture events and stories, enabling us to explore the editorial aspects of the approach.
Some of the things we’ve learned are substantial. The range of situations described in the FrameNet semantic database seems to be sufficient to cover the majority of news events that we are seeking to report, which is very encouraging. A significant proportion of reporting seems to involve speech acts and other forms of communication by characters – probably to an extent that will require special handling of those kinds of events. There seems to be a previously unappreciated challenge in distinguishing between the structuring of events from language and the structuring of events from ‘models’ of stories. There are several built-in trade-offs in the nature of the event frames that we are creating – for example general vs specific, or across multiple FrameNet frames – which will probably require a shallow taxonomy of event frames (as FrameNet itself already has). Reporting and editing tools for handling characters and entities that have no external knowledge graph references will need to be substantially improved.
We have also learned much that will enable the development of nascent editorial guidelines to aid future structured reporting – how to define and choose event frames, how to choose between importance values and sub-narratives to represent detail, how to name characters and entities, how to systematically select external references for characters and entities, how to approach the specificity required for capturing structure. The list of software issues to be fixed and improved is also long and somewhat daunting. We have not yet come across any specific issue that suggests an insurmountable editorial barrier to the concept, although there are still lots of puzzles, questions, weirdness, vagueness and things-to-explore that may yet prove to be major challenges.
This isn’t easy. We are attempting to record general news events and news stories as structured data, which is a radical and unexplored notion. Success of any kind is not guaranteed and the events and stories that we are reporting and recording may be somewhat simplistic, coarse and clunky. All of this is, obviously, much harder than just writing more text. But we are actually reporting and recording general news as structured data. That is actually happening. For real.
June 8, 2015
In the month since my last blog post I’ve made three week-long trips across the country, engaging with the two communities most closely associated with Structured Stories – Journalism and Computational Narrative.
My first trip was to the Reynolds Journalism Institute at the University of Missouri in Columbia, where I am a fellow this year, conducting a formal evaluation of Structured Stories. I met most of the RJI leadership team, including Executive Director Randy Picht and Research Director Esther Thorson, and spent several days with my research partner at Mizzou, Frank Russell. I was impressed by the intellect and seriousness of everyone I met and I’m convinced that RJI is the perfect environment for a careful, thoughtful and credible evaluation of the concept. I will publish more details on our research program as Frank and I develop the particulars.
Trip two was to Atlanta, Georgia, where the 6th workshop on Computational Models of Narrative was taking place. I was there to deliver my paper on ‘Narrative Structures as a Framework for Journalism”, and this was the first time I had presented the Structured Stories concept to the computational narrative community. I was very pleased by the interest and response, and I came away feeling confident about the conceptual basis of the approach and about the place of Structured Stories within the field. I also met many fascinating people with long experience in representing narrative as data, made new friends at the workshop and over dinner each night, and was introduced to several other people doing interesting and related work in the Atlanta area.
The third trip was to New York, New York – specifically the exciting and fast-paced neighbourhoods of Soho and TriBeCa. This trip was for the training program for the team participating in the Structured Stories NYC reporting project. The team members are Ishan Thakore, Natalie Richie and Rachel Chason – all students from Duke recruited and guided by Bill Adair. We were also joined by several guests, and went from an introductory overview to structuring real events and stories in three days. It was an intense experience filled with interesting discussions and examples, and the high calibre of our reporters made me very pleased with how the project has kicked off. Structured events from the NYC project are already pouring in and stories should be up on the website within a few days.
I am now back in L.A. for at least the next 2 months, focused primarily on supporting the reporting team in NYC. The NYC project is critical because it will determine whether the Structured Stories concept is editorially feasible – i.e. can it work on real stories in a real reporting workflow. With this project we are exploring ‘structured editorial’ issues that are new for journalism, and we may uncover many unanticipated challenges and opportunities. These are still very early days for Structured Stories, but they are increasingly busy and filled with interesting engagement!
May 5, 2015
More good news!
I am very happy to announce that I have been selected as a 2015/16 Fellow at the Reynolds Journalism Institute, focused on the evaluation of structured narratives as a new framework for journalism using the Structured Stories platform. The fellowship will give me access to the deep academic and practical expertise of RJI and the wider Missouri School of Journalism community, and will help ensure that structured narrative gets wider exposure and a thorough evaluation as a possible new medium for journalism.
The timing for the fellowship could not be better. Reporting into the Structured Stories platform will begin soon, and evaluating news consumption using the accumulated reporting will be greatly aided by the involvement of experienced journalism researchers.
The RJI fellowship is non-residential. I will remain in Los Angeles for the duration and will travel frequently to RJI in Columbia, Missouri.
April 24, 2015
New York City, here we come!
I am extremely pleased to announce that a major structured journalism reporting project, using the Structured Stories platform and titled ‘Structured Stories NYC’, will take place in New York City this summer. The project is a partnership between the Reporter’s Lab at Duke University, the WNYC newsroom in New York and Structured Stories. Reporting will be done by a team of student reporters from Duke organized by Bill Adair – the new Director of the DeWitt Wallace Center for Media and Democracy at Duke University, creator of PolitiFact and Pulitzer prize winner – and the reporting team will be led by Ishan Thakore from Duke.
The Structured Stories NYC project has been made possible by a generous grant from the Online News Association’s Challenge Fund, which was announced earlier today at the Journalism Interactive 2015 conference.
This project marks a significant milestone in the progress of Structured Stories, because it will provide an opportunity to develop real-world editorial guidelines and processes for structured journalism, and will result in a substantial body of reporting in structured form. The Structured Stories approach to local government journalism may offer a new way for citizens to quickly understand large, long-term, sprawling local government stories – the kind of stories that would otherwise need to be followed closely over a long time to be deeply understood.
A detailed description of the Structured Stories NYC project can be found on the ONA Challenge Fund website here.
Also, on a more technical note, my paper titled “Structured Narratives as a Framework for Journalism: A Work in Progress”, has been accepted for the 6th Computational Models of Narrative workshop (CMN’15), which will be meeting at Georgia Tech in Atlanta on May 26-28. The paper is a technical description of the Structured Stories technology and data structures, and of the journalistic basis of the approach. I will be presenting the paper at CMN’15 and will link to it here after it has been published.
These are very good developments for the Structured Stories project, and offer opportunities to truly explore and evaluate structured narrative as an alternative approach to journalism. If you are interested or intrigued then please get in touch!
You can reach me on email at david[at]structuredstories.com and on Twitter at @StructStories.
March 11, 2015
The past 6 weeks have been mostly consumed with activity that was important to moving Structured Stories forward, but was not coding and was not reporting. More will be forthcoming in the fullness of time.
The bit of coding that was done in February was mostly about building a nascent editing tool, which isn’t publicly visible but which will enable publicly visible things to happen – namely many more events and stories.
The editing tool will do a lot of things (see the previous post), but at the moment the one big thing that it needs to do is enable the easy creation and editing of event frames. Event frames are at the heart of structured stories and working with them quickly and easily is essential to establishing a reporting process.
Nothing is easy, and everything takes more time that hoped, but things are moving forward. My goal for 2015 is to enable reporting into Structured Stories sufficient to enable a robust evaluation of the concept. There are many aspects to that goal that will require more exploration, trial, error and time. If you want to help then get in touch.
December 31, 2014
With just a few hours left in 2014 I figured it was time for quick recap of the year that was and a quick preview of the year to come. I also just pushed a major update to the ‘production beta’ code, including the ability to use Facebook reference IDs to define characters.
A year ago Structured Stories was just a goal, a pile of research notes, some nascent ideas and a primitive prototype. The data model and architecture stabilized in Q1 and design and coding of the application began in Q2. The early beta application launched in October and has been continuously improved since then. Today it is an increasingly stable tool with an accessible user experience and a robust API, and users can use it to consume and create structured events and structured stories. Judge for yourself here.
2015 is about transitioning the Structured Stories project from coding to journalism. The beta application will be function-complete within a few weeks, and by February I hope to have editing tools in place sufficient to enable user-entered stories to become permanent. At that point the project changes from a technical focus to a journalism focus – coding will scale back to mostly bug fixes, and the creation and growing of stories will become the primary activity. Local government news in Los Angeles remains the domain.
Focusing on journalism requires discovering, understanding and addressing a complex set of editorial challenges that will probably be at least as daunting as the technical challenges of the past year, including:
- Creating and applying editorial guidelines for the creation of event frames.
- Creating and applying editorial guidelines for the entry of events and the creation of stories.
- Building the event frame library from just over 100 frames now to several thousand frames.
- Developing an editorial process that can accommodate many contributors to stories and that can support coherent editing of events and stories.
- Understanding how story and event editing and maintenance actually work and building tools to support those activities.
- Observing and reacting to how real users use Structured Stories to create and consume stories.
- Discovering how to educate early users about Structured Stories, its functionality and utility.
The primary usefulness of the beta application in 2015 is to enable these editorial challenges and others to be clearly identified, defined and addressed. There is much to do and much to learn, but there has also been some progress. The Structured Stories concept works technically, and if it can also work editorially then it may be useful.
November 7, 2014
The Structured Stories beta application launched publicly in late October, 2014, after a long and painful delivery.
The application is now open for anyone to browse and explore structured stories, but registration – required for creating and editing stories – is still ‘upon request’ until at least the end of 2014. Pick a story and look at the ‘structured story’ view to quickly understand the concept.
This is a milestone, but these are still very early days for Structured Stories. The next few months will be about debugging and completing the full v1.0 stack (backend apps, API and web app), and then the focus will shift from technology to journalism. The rough goal is to build out a dense collection of stories (a ‘narrative network’) on local government in Los Angeles by mid-2015 – sufficient to fully demonstrate the concept, test its value and generate learnings about the application of semantic and knowledge engineering techniques to news.
At this stage of the project FEEDBACK IS CRITICAL. If you have opinions or criticism or comments about this project, or if you find bugs, then please let me know.
September 22, 2014
The first releasable version of Structured Stories is almost ready and will launch in October in five languages (English, Spanish, Swedish, Irish Gaelic and Indonesian), focused on local government news in Los Angeles. Yes, you read that correctly.
I have been demonstrating the application in Los Angeles since August and will demonstrate it to interested people at the Online News Association conference (ONA14) in Chicago in late September. I invite anyone who would like to see it to either meet me at ONA for a live demo or to get in touch for a remote demo (my email address is david[AT]structuredstories.com). In the meantime I am including a short series of screenshots below to convey the gist of the application.
After many months of immersion in the design and creation of the public-facing Structured Stories application I believe that there are at least six major novel functions that this approach can provide to news consumers:
- It makes news permanent. By enabling news to accumulate over time in a way that can be consumed intuitively and naturally, Structured Stories turns news from flow into stock – building a permanent history from news streams.
- It enables highly efficient consumption of news stories. By organizing news as narrative structures instead of as written text, Structured Stories makes it simple to navigate and understand vast and sprawling news stories that would otherwise be inaccessible without significant research. Navigating single stories at the scale of libraries becomes possible.
- It makes news universal. Structured Stories does not use language as its primary representation mechanism for news, and therefore any news can be easily consumed in any language. The prospect of a single, global news platform that is equally accessible to anyone in almost any language becomes a very real possibility.
- It enables queries of news and reasoning on news. Because Structured Stories are, well, structured, they are accessible to explicit search queries (think SQL, not Google) and are also available to computational reasoners. The application of computational tools like machine learning to all/any news events is realistic.
- It separates the method of storing news from the method of consuming news. There are many, many ways to tell the same story, and Structured Stories enables all of them, using built-in features, custom ‘discourse elements’ and through a Structured Stories API that enables unique story readers, story viewers or other story display concepts. Even video is possible.
- It publishes journalistic news events as Linked Open Data. By providing each individual news event, no matter how small, with its own unique address on the Internet, Structured Stories can open up news sharing and news mashups in lots of new and exciting ways – facilitating entirely new forms of discussion, verification and validation.
The Structured Stories approach to news also has characteristics that enable multiple novel functions for news producers. By separating reporting from writing it becomes practical as an adjunct to existing newsroom processes – a typical local government newspaper story of 3-4 core events can be entered into Structured Stories in under a minute. At the same time it also enables the possibility of ‘re-bundling’ news as value-accumulating networks and has the potential to reconcile civic journalism with professional journalism and editorial oversight in an economically sustainable way.
A demonstration is highly recommended.
Screenshots of the Structured Stories BETA application:
A detailed description of the Structured Stories technology in PDF form is available here.
August 4, 2014
Well – so much for ‘updates coming soon’! I have been so consumed in technical matters for the past two months that I haven’t had time to do much else. The explanatory presentation remains unfinished and I’m beginning to feel guilty about not keeping my small handful of blog readers informed. So here’s a quick post to check in and to give a (hopefully) more realistic estimate of what to expect next and when to expect it.
The beta product is generally on track and is dramatically simpler and easier to grasp than the prototype. The functionality is almost identical to the prototype – no more and no less – but the user interaction is unrecognizable and (I think) a vast improvement. I am aiming to open the beta site to anyone who is interested in the last week of September, following the ONA 14 conference in Chicago. I will be attending ONA14 and demonstrating the beta application there, so if you would like to meet up please get in touch. More details to follow. Really.
The central update from the last few months is that Structured Stories is advancing from being just a technology to being a real product – a process that has been rooted in many iterations of the user experience design and that includes deep product decisions that have been forced by the UI development (most prominently the decision to include reporting tools directly in the product from the get-go). I have already begun to demo the beta application and we should be only about 7-8 weeks away from a product that YOU, dear blog reader, will be able to play around with yourself – browsing and entering stories. This is hard. Please be patient.
As always, I invite anyone who might be interested in Structured Stories to get in touch. I always have time to chat or email and I can even set up a web-ex for a sneak preview of the beta. Have a great summer and look forward to playing with a new approach to news in the fall!
May 21, 2014
The New York Times internal report on innovation, leaked last week in the aftermath of Jill Abramson’s exit, has set news innovation circles abuzz and is filled with many observations and recommendations that argue for more structure-centric journalism – including structure based on story. The business and product case for news producers to extend the ‘value half-life’ of their journalism by adding structure certainly seems to be building rapidly, and for a very comprehensive treatment of that phenomena I highly recommend Reg Chua’s Structured Journalism blog.
Structured Stories fits well with this increasing attention to structure in news production. The technology can be interpreted as a generalized mechanism for structuring news, enabling the practical application of structure beyond one-off domains like homicides, recipes, etc. to include news of any kind, in any domain. Such a generalized mechanism for fully structuring news, if productized and if successful, could be economically positive for news producers by somewhat restoring the old ‘news bundle’ in the form of interconnected and proprietary data.
It is important to keep in mind, however, that news production is just one side of the news ecosystem and that massive challenges also exist in the consumption of news. These challenges have been well described by Tony Haile, the C.E.O. of Chartbeat, in a recent article informed by Chartbeat’s unique view into the click-by-click realities of digital media consumption – realities that include 55% of clicks resulting in less than 15 seconds of user attention, and that show no discernible relationship between the sharing of content and user attention to that content. This describes a media ecosystem in which news consumers are expected to act as their own ‘do-it-yourself’ content editors; to pick through, assess and accept or reject content, one article at a time, every day. This is, as software developers say, ‘suboptimal’.
Structured Stories can help here too. By converting news into a ‘permanent record’, accessible on the consumer’s own terms and timeframe, and by providing a naturally intuitive framework by which news can be accessed, navigated and queried, Structured Stories can provide consumers with a new editorial structure for news. Replacing the text article with the structured narrative as the primary ‘unit of news’ would ground every news event as a URI on the semantic web, and therefore would provide a permanence and a degree of interconnection that is inconceivable within the current article-centric news ecosystem. Structure adds editorial value to information, and structures stories can add editorial value to news.
Update: I am currently heads-down on API design and architecture, and I hope to post a detailed technical description of that in the next few weeks. Also, I received feedback suggesting that an overview presentation describing the Structured Stories concept would be useful in interpreting the demo, and I should have that available soon.
April 9, 2014
Since February I have been doing some outreach about StructuredStories and have come to realise that I need to communicate more clearly about the technology. I am therefore providing two related guides – one in the form of a video demonstration of the prototype, and the other in the form of an informal FAQ document.
The video demonstration is quite extensive – about 1 hour – and my goal has been to replicate what an interested layperson would experience if I were demonstrating the prototype to them personally. I provide lots of commentary, including brief introductions to knowledge graphs and to FrameNet, and I apologise in advance for butchering some of these descriptions in attempts at simpler explanations. The video is on YouTube – I recommend that you view it in High Definition mode. The link is here:
[IMPORTANT!: The link above is a video demo of the *proof-of-concept prototype*, not of the beta product, and it is intended for those interested in the technical background behind Structured Stories. If you want a demo of the beta product then either get in touch or wait until it launches in mid-October, 2014.]
The Frequently Asked Questions document is loosely based on questions that I receive while demonstrating and talking about the technology, and I intended for this document to accompany the demonstration. I suggest that you read this document first, then decide if you want to view the 1 hour demo. The FAQ document is in PDF format and is linked here:
The StructuredStories project is now in a new phase. I have decided to focus my attention on the specification and architectural design of a ‘version 1’ product, and therefore to place less emphasis on the prototype except for experimentation and communication. This will increase the time until I can release a publicly-accessible beta product, but will hopefully result in the creation of an enduring suite of technology. The prototype has largely served its purpose. It works.
February 25, 2014
Well, it has been a very busy two months! January was very challenging but extremely productive and was centered on two additional major revisions to the event and narrative data structures. This brings the number of major model revisions to five – this process of alternatively coding a working prototype and then assessing the resulting functionality has been absolutely critical to moving the Structured Stories project forward, as has a deeper review of various research literatures in light of learnings from the prototypes. The system is now very much clarified, and I have spent most of February documenting it and preparing the patent application. That application has now been submitted, and therefore I can now be a bit more forthcoming here about the details. This will be a highly technical description, but I am publishing it here anyway in an effort to be as transparent as I can about this project. Please email me if you would like more (or less!) details.
The basis of the Structured Stories system is the concept of ‘semantic events’, which are uniquely identifiable representations of any specific activity in the world, no matter how large or small. These semantic events are defined using a library of abstract definitions of general forms of semantic events, which are in turn defined by a formal library of semantic units (specifically the FrameNet semantic library from UC Berkeley). The interior definitions of these semantic events are based on two basic elements – (1) a set of formal semantic roles describing the specific activity of the characters, entities, locations, etc that are involved the event and which are ‘filled’ using typed references to FreeBase, WikiData, GeoNames, etc, and (2) semantic phrases, primarily verb-based predicate phrases, that convey the activity that the event is about and which are semi-formal. The use of these partially-formal definitions of events, based on an open-ended library of abstract event forms and grounded in FrameNet, is intended to enable any conceivable event to be captured and represented without primary dependence on natural language. Furthermore, the availability of uniquely identified and partially-formal semantic events enables quite a few additional features at the event level, some of which are extremely powerful – for example the ability to represent and use cause-and-effect relationships between discrete events.
But capturing and representing semantic events is only half of the challenge here. The other half is in providing a method for organizing and navigating those events in a manner that is not merely coherent, but is also optimized for human understanding of the underlying events within their context. This method is the ‘narrative structure’, or ‘structured story’. A narrative structure is set of references to semantic events that enables those events to be ‘consumed’ as a narrative, and is much more than merely a simple list of events. Narrative structures include recursive elements (or ‘sub-narratives’), importance weightings and mechanisms for navigation between stories via common events, all of which mirror the natural narrative capacity of human beings.
There are a lot of interesting things that result from this representation of semantic events and of narrative structures. The most significant is the establishment of an ‘event’ graph’ and/or a ‘narrative graph’ that is formed from the various relationships between events and narratives – it has been interesting to observe this event/narrative graph emerge even from just the 40 or so L.A. local government stories that have been represented within the system to date. The power of this representation for extremely specific question answering has also become very apparent, as has its potential as a powerful way to navigate the ‘document web’ based on narrative.
December 23, 2013
As 2013 wraps up I have been reviewing progress since beginning the Structured Stories project some 5 months ago. The key question at this stage is whether the Structured Stories technology is currently delivering an experience of news that is genuinely different from existing digital news channels. I believe that it is, although whether others will agree with me, or whether that experience is different in a useful way, or whether that difference can feasibly be delivered at scale are all questions that can only be answered by releasing a public newsreader and then gathering and assessing analytics from actual use.
Nonetheless I think it might be useful to describe this different experience of news as perceived by a few early testers using the development UI. I cannot publish full details about the technology or about the experience that it delivers until I have some intellectual property protection in place, but I can give an overview of the general texture of the experience.
Keep in mind that the events and causal relationships that are currently stored in the prototype narrative structures are solely within the domain of Los Angeles city government during November and December 2013. Quantitatively this represents approximately 25 stories, most of which contain between 5 and 15 events – an event density that is very roughly equivalent to about 80 ‘traditional’ news articles about L.A. local government. The capture/reporting of these events is also very elementary and is done from mainstream press reports, from a daily summary document provided by the L.A. City Clerk’s office and from blog posts and press releases from key characters in the stories (primarily L.A. city council members).
Given all these caveats, my major observations are as follows:
1) Efficiency of news consumption. Readers seem to be able to consume a lot of ‘real news’ in a short time. In terms of ‘number-of-events-per-minute’ the improvement seems to be substantial – possibly 5x or more. Furthermore, this efficiency seems to be achievable with improved comprehension, as measured by an informal assessment of ability to answer key questions about the narrative. More systematic measurement of the efficiency of news consumption on a production environment is possible via formal user testing and will provide definitive metrics on this.
2) Integrated path to detail. The effect of integrated access to ancillary information or content relevant to either individual events or to narratives is quite striking. Exploring within the Structured Stories environment (e.g. quotes or other discourse elements) is useful, but it is the effortless direct access to raw sources, to knowledge graph or wikipedia entries, to a variety of discourses (articles, etc) about particular events/narratives and to other related linked information that is most powerful. Furthermore, accessing external resources from the Structured Stories UI does not feel like being ‘interrupted’ while consuming a narrative via a text article, but instead feels naturally integrated into the experience of consuming the narrative. It is not merely that this ability to access ancillary information is interesting, but it also enables one to feel in full control of the interpretation of events and narratives. Some measurement of this phenomena is probably achievable with relatively standard analytics.
3) A sense of coherence. Subjectively, consuming news via the Structured Stories environment just feels right. It feels complete, authentic and efficient – one evaluator described it as combining the ease of bullet points with the depth of long-form articles. Part of this probably comes from reader’s natural affinity for clean narrative structure, but the experience of coherence is also likely delivered through the clear visibility of events, of their relationship to other events and of their place within the overall narrative – all of which make it easy to continually ‘know where you stand’ while consuming a narrative. I am seeking a quantifiable metric for this sense of coherence, and I suspect that it may be the key enabler of the seemingly improved comprehension of narratives mentioned earlier.
4) A sense of control. The experience of consuming news within the Structured Stories prototype environment is primarily an experience of control. You are constantly making decisions about what to pursue, how much detail you require, how much ‘color’ you require, whether you need to or want to read external discourses/articles, etc. There is no sense of having to invest in consuming part of a story before knowing whether it is interesting, or of being passively fed a stream of content morsels, or of a mismatch between your required level of detail and the level of detail in the presentation, or of suspicion about the value or credibility or relevance of either content summaries or the content itself, or of regret after investing in consuming a story. It is not a search experience, because the entire experience is guided by the narrative structure, but it is ‘search-like’ in its level of control and engagement – for better or for worse.
5) Permanence. It is probably a little early to be commenting on the ‘sense of permanence’ that is engendered by the Structured Stories environment, but I think that it is important. Although completely focused on news, the Structured Stories environment is engaged with as ‘stock’ rather than as ‘flow’. Some events in the Structured Stories narratives date back years, because they are the causes of current events. Other events arise and are added to existing narratives, even thought those narratives may have already seemed ‘complete’. There are far fewer stories than articles and these stories and their constituent events are permanent artifacts – much more like Wikipedia than like digital media news streams, even for same-day breaking news. This changes how they are perceived, especially when engaging with the same story as it develops over days or weeks or, eventually, years.
There are other potentially valuable characteristics of the Structured Stories approach that are not directly related to user experience, such as its compatibility with mobile and cross-platform consumption, its generation of rich analytics and the natural suitability of narratives for sharing on social media, etc. All of that, however, is irrelevant unless the deep experience of using this approach is genuinely different – and these very early, subjective and informal observations suggest that it might be. I hope to open the demonstration site to the public by April 2014 so that anyone who is interested can judge for themselves.
December 5, 2013
Working with stories in computers is obviously dependant upon a knowledge representation mechanism that allows for the storage and retrieval of narrative information – a narrative data model. The Structured Stories narrative data model is now in its third major revision, and expresses all of the major requirements for working with stories within the domain chosen for the proof-of-concept project: local government news. The model has been influenced by published models used in various pioneering computational narrative projects, but has some fundamental differences from these – most importantly in the use of semantic web concepts and in its focus on micro-content digital media requirements rather than on literary or academic objectives. Although it is likely that the model will undergo a further two or three major revisions before it ready to support a beta release, it is now beginning to stabilize and it currently supports the addition of new events and new stories at a rate that is equivalent to existing news reporting in the domain. Revisions of the model are primarily driven by learnings from working with these ‘live’ events and stories, and I expect to publish more details here once the model has stabilized and I have filed key patent applications.
I have found that attempting to capture the abstraction of stories in a useful way is not easy, however after months of modeling and three major revisions I have not yet come across an insurmountable conceptual obstacle. I believe that my narrative data model is growing increasingly useful – at the very least within the domain of local government news.
November 6, 2013
It’s not news that the news industry is experiencing a period of remarkably intense change, driven by the rise of digital technology and the resulting demise of barriers-to-entry and bundled media products. There is no shortage of commentary, analysis and discussion about how this change is affecting the production and distribution of journalism – for example in the excellent report by C.W. Anderson, Emily Bell and Clay Shirky titled “Post Industrial Journalism”, published recently by the Tow Center for Digital Journalism at Columbia University. These are indeed interesting times for the production and distribution of news.
We see, however, relatively little analysis or discussion of how these changes are affecting the consumption of journalism by individuals. Questions that seem central to the very purpose of journalism remain either intuitive or unasked: Why do people even consume news? What are the fundamental elements of news, as experienced by the individual? What is the best way to organize news in terms of the individual needs that it addresses? How do the reasons for consuming news vary by individual, by context, by subject matter, etc? How well does the news environment satisfy those reasons? Are those reasons changing? The lack of obvious interest in these questions is particularly surprising given the vigorous search for innovative ways to offer value in the digital media environment, and suggests that it has been difficult to identify or articulate a ‘theory of news’ that could serve as a starting point for exploring the consumption side of the new marketplace for news.
It should come as no surprise that the hypothesis that I propose as the basis of a possible ‘theory of news’ is centered on the critical role of narrative, or story, in people’s perception of their world. While it is obvious that news is about ‘stories’ in the intuitive or colloquial sense, it is perhaps less obvious that the growing body of cross-disciplinary work on narrative can be directly and formally applied to understanding the consumption side of digital media or that such an understanding of news consumption might be used to produce useful media products.
Take, for example, an experience that is probably familiar to anyone who consumes news from many different online sources (news sites, Twitter, Facebook, personalized streams, news apps, blogs, etc.). This behavior can be interpreted as being about seeking new stories that are interesting to that consumer and seeking new developments in interesting stories that the consumer is already aware of. As the consumer develops an interest in a particular story they must therefore personally ‘de-duplicate’ events across multiple documents in order to construct the full story in their mind, and their personal experience in doing this becomes something like ‘knew that’ – ‘knew that’ – ‘knew that’ – ‘thats new!’ -‘knew that’ – ‘knew that’, etc. They are forced, in effect, to act as their own editors – a new role for consumers that was much less necessary in the pre-digital media environment. This frustrating experience is exacerbated by the very tools provided by digital media sites in an effort to simplify the discovery of interesting content, such as document-oriented topic clustering, entity-based personalization of news streams or search tools. From the consumer’s perspective navigation and consumption based on stories and their constituent events rather than on documents would probably be more satisfying.
October 21, 2013
Welcome! This blog is a key component of the StructuredStories project and is the heart of this website. I will use it to share successes, challenges and detours along the path towards narrative-based applications in digital media, as well as posting a few book reviews, technology assessments or commentaries from time to time. Connecting with others with similar interests is much of what makes these endeavors worthwhile, and so I hope that actively and transparently reporting on my progress will help to facilitate that.
I have been keenly interested in computational narrative for several years, and interested in the digitally-driven changes in journalism and media for much longer than that. I began to realize that these interests might overlap, and furthermore that several rapidly maturing technologies might make it possible to realize useful applications within that overlap. My work at Yahoo! took me deeply into digital media use cases and systems (especially automated content understanding and personalization), but I felt that narrative-based approaches to media were as yet too undefined to reasonably pursue within a corporate environment. I left Yahoo! in July 2013 and began to systematically assemble a research and prototyping program aimed at developing proof-of-concept digital media applications based on narratives. As I write this post I am now three months into this program, and my excitement about the possibilities from this approach is growing.
StructuredStories is a company but not yet a startup. This sounds confusing, but is just an expression of how I like to work on complex product design problems. I like to reduce large and gnarly problems to first principles, and then to build new approaches to those problems from the ground up – thinking through, writing down, drawing out, coding up and trying out thoughtful solutions. This approach requires lots and lots of time and intellectual freedom, and is somewhat cross-threaded with the rapid ‘code-test-learn-repeat’ approach to product development that is currently popular. While the risks of ending up in a product ‘dead-end’ are higher with this approach, it also offers the potential for finding the ‘global maximums’ of a product solution space, rather than just the ‘local maximums’ available from more evolutionary approaches. StructuredStories is therefore a vehicle for me to pursue a ‘global maximum’ in digital media products based on narratives, while working in the way that I like to work.
Its hard to imagine anything as simple to the human mind as a story, but applying analytical and computational narrative to digital media is nonetheless insanely complicated. A first challenge is just in thinking clearly about narrative – something that is simultaneously highly abstract but also as familiar as breathing. A second challenge is in integrating what I call the ‘large number of small literatures’ representing the contributions of about a dozen different research fields to the study of narratives in various guises - a challenge that has provided me with several years of enjoyable reading, dot-connecting and occasional but satisfying glimpses of recognition. A third challenge has been developing a working understanding of the foundational issues facing digital media from both the production and consumption sides – issues like the drivers towards media personalization, the changing role of editorial influence in media, the deep causes and effects of habit-driven media and the biases within the existing digital media establishment. A fourth challenge is, of course, understanding a range of new technologies, matching them to the pragmatic requirements of platforms and applications based on structured narratives and then building useful prototypes.
At this stage I make no guarantees about whether this approach to digital media is either technically feasible or commercially viable, however I strongly suspect that it is. Either way I intend to find out, and I invite anyone who is interested in the question to follow this blog and to get in touch.