Natural Language Understanding requires a large amount of background "common sense" knowledge about the situation under discussion. In many respects, using this knowledge is at the core of reasoning and acting in traditional Artificial Intelligence. When reading an article about a criminal conviction, the writer assumes the reader knows about trials, juries, and criminal activity. The Narrative Chain project aims to learn this knowledge by processing large amounts of text and learning which events tend to occur together. We are studying not just what can be learned, but also the best representation for this knowledge (graph, linear chain, frame?).

This project also includes research into ordering events in time. For instance, did the conviction or the sentencing happen first? We use modern machine learning techniques to find linguistic features that indicate this semantic ordering relation.

An example of a learned narrative event chain, with arrows indicating temporal ordering, is shown on the right. The bold words are the events, and the subj/obj terms indicate how the common actor in this narrative is involved in the event (the subject or object of the verb).

The obvious question is how one manages to distinguish among summary, analysis, argument, and interpretation. E.g.

With the aid of a rag tag crew of adventurers, a young man rescues a princess from an evil empire and discovers his destiny to become a member of a dying order of knights.

A young man helps a rebel leader escape from an imperial prison and participates in an pitched battle to save the rebels' military base.

I assume you recognize the story, and I think most people would say the first summary is more accurate. Why? The second one is certainly not inaccurate. It simply downplays the "hero's journey" aspect and portrays the film as depicting a political and collective activity.

Temporal Clustering (TC) refers to the factorization of multiple time series into a set of non-overlapping segments that belong to k temporal clusters. Existing methods based on extensions of generative models such as k -means or Switching Linear Dynamical Systems (SLDS) often lead to intractable inference and lack a mechanism for feature selection, critical when dealing with high dimensional data. To overcome these limitations, this paper proposes Maximum Margin Temporal Clustering (MMTC). MMTC simultaneously determines the start and the end of each segment, while learning a multi-class Support Vector Machine (SVM) to discriminate among temporal clusters. MMTC extends Maximum Margin Clustering in two ways: first, it incorporates the notion of TC, and second, it introduces additional constraints to achieve better balance between clusters. Experiments on clustering human actions and bee dancing motions illustrate the benefits of our approach compared to state-of-the-art methods.

When the Palestinians launch a mortar attack into Israel, the Israeli army does not wait until the end of the calendar year to react. Yet, most modern data collections are aggregated to the month or year. The data available here include almost 10 million individual events, each coded to the exact day they occur or become known. Each event is summarized in the data as "Actor A does something to Actor B", with Actors A and B recording about 450 countries and other (within-country) actors and "does something to" coded in an ontology of about 200 types of actions. The data are coded by computer from millions of Reuters news reports. The software system (produced by VRA) that performs this task has been independently evaluated by King and Lowe (2003). This article found that for the numbers of events it was possible to convince humans (trained Harvard undergraduates) to code by hand, the machine did as well as the humans. For much larger numbers of events for which no expert coder could keep up, the machine dominates.

We present a working system for large scale quantitative narrative analysis (QNA) of news corpora, which includes various recent ideas from text mining and pattern analysis in order to solve a problem arising in computational social sciences. The task is that of identifying the key actors in a body of news, and the actions they perform, so that further analysis can be carried out. This step is normally performed by hand and is very labour intensive. We then characterise the actors by: studying their position in the overall network of actors and actions; studying the time series associated with some of their properties; generating scatter plots describing the subject/object bias of each actor; and investigating the types of actions each actor is most associated with. The system is demonstrated on a set of 100,000 articles about crime appeared on the New York Times between 1987 and 2007. As an example, we nd that Men were most commonly responsible for crimes against the person, while Women and Children were most often victims of those crimes.

Event coreference is an important and complicated task in cascaded event template extraction and other natural language processing tasks. Despite its importance, it was merely discussed in previous studies. In this paper, we present a globally optimized coreference resolution system dedicated to various sophisticated event coreference phenomena. Seven resolvers for both event and object coreference cases are utilized, which include three new resolvers for event coreference resolution. Three enhancements are further proposed at both mention pair detection and chain formation levels. First, the object coreference resolvers are used to effectively reduce the false positive cases for event coreference. Second, A revised instance selection scheme is proposed to improve link level mention-pair model performances. Last but not least, an efficient and globally optimized graph partitioning model is employed for coreference chain formation using spectral partitioning which allows the incorporation of pronoun coreference information. The three techniques contribute to a significant improvement of 8.54% in B 3 F-score for event coreference resolution on OntoNotes 2.0 corpus.

Abstract. This paper proposes a corpus-based study of how texts guide readers through time. It focuses on sentence-initial temporal expressions which, beyond locating an event in time, take on a discourse dimension via the process of “indexing”: contrary to connection, which looks backward towards previous text, indexing, also referred to as “discourse framing”, looks forward and provides instructions for the interpretation of forthcoming text. As a first step towards an investigation of the impact of genre on temporal framing, our French language corpus is constructed according to the most crucial distinction regarding temporality: narrative/non-narrative. The non-narrative sub-corpus provides archetypal examples of text organisation through temporal framing. Narrative texts on the other hand, because of the interaction between framing and another major mode of temporal organisation – through the Narration relation – resist the indexing model and therefore force us to refine the notions.

In two experiments we assessed the degree to which memory for events are similar or differ depending on whether they were narrative or autobiographical events. Consistent with previous research on autobiographical memory, memories for events captured the sequential order of events. However, in contrast to autobiographical memory research, ratings of importance did not appear to be related to retrieval speed. An analysis of causal connectivity of the recalled events was significantly related to retrieval speed. Issues of narrative comprehension and memory, autobiographical memory, and their overlap are discussed.

This study examined the forgetting curves for information read in a novel. People read a 10-chapter novel where each chapter covered an approximately 10-year period in the life of the protagonist. After reading the entire novel, participants completed various memory tests in which they summarised the novel, provided associated information from cues, and answered specific questions. Performance was plotted as the amount of information or the accuracy of question answering for each chapter. All of the memory tests revealed similar patterns: (a) better performance for early information (a primacy effect), (b) a bump in performance when the protagonist was approximately 20 years old, and (c) a smaller bump in performance when the protagonist began a career later in life. These results are considered in the context of theories of forgetting, autobiographical memory, and situation models.

There are also a number of potential applications to information technology. Interfaces designed to teach procedures or scientific processes may benefit from explicitly representing the event structure of the activity for the learner (Zacks & Tversky, 2003). Psychologically adaptive segmentation may provide an efficient way of summarizing large databases of video or multimedia for search and editing (Christoffersen, Woods, & Blike, 2007). Identifying event boundaries may be helpful in scheduling interruptions in the context of tasks such as piloting, driving, or operating machinery.
Finally, event segmentation may provide a powerful lens through which to view art and literature. One important thing that cinema, television, and literature do is represent events. Some basic features of these ubiquitous media are still poorly understood. For example, how is it possible that a film can cut from one time and place to another, instantaneously changing all the information in the visual field, without disorienting the viewer (Münsterberg & Griffith, 1916/1970)? One possibility is that the perception of events regulates how cuts are perceived and which sorts of cuts “work” (Zacks & Magliano, in press). What does a reader retain over the reading of an extended novel (Copeland, Radvansky, & Goodwin, 2009; Radvansky, Copeland, & Zwaan, 2005)? The behavioral and neurophysiological data suggest that readers construct event representations that are segmented according to the same mechanisms as govern the segmentation of live action (Speer et al., 2009; Zacks et al., 2009). Thus, the chunking of experience into events may enable disparate artistic forms to convey experience.

When reading a story or watching a film, comprehenders construct a series of representations in order to understand the events depicted. Discourse comprehension theories and a recent theory of perceptual event segmentation both suggest that comprehenders monitor situational features such as characters’ goals, to update these representations at natural boundaries in activity. However, the converging predictions of these theories had previously not been tested directly. Two studies provided evidence that changes in situational features such as characters, their locations, their interactions with objects, and their goals are related to the segmentation of events in both narrative texts and films. A 3rd study indicated that clauses with event boundaries are read more slowly than are other clauses and that changes in situational features partially mediate this relation. A final study suggested that the predictability of incoming information influences reading rate and possibly event segmentation. Taken together, these results suggest that processing situational changes during comprehension is an important determinant of how one segments ongoing activity into events and that this segmentation is related to the control of processing during reading.

The War of 1812 has long been a conundrum for school children and historians.  Fought due to miscommunication and to a large extent between irregular mobs, the War of 1812 has one final and irrepressible problem above all, the name does not successfully describe the conflict.

In recent years, researchers in several communities involved in aspects of the web have begun to realise the potential benefits of assigning an important role to events in the representation and organisation of knowledge and media. While a good deal of relevant research has been done in the semantic web community (for example on the modeling of events), a lot of complementary research has been done in other communities, such as multimedia processing and information retrieval. The goal of this workshop is to advance research on this general topic within the semantic web community, by both building on existing semantic web work and integrating results and methods from other areas, with a particular focus on issues that are central to the semantic web.

With the proliferation of citizen reporting, smart mobile devices, and social media, an increasing number of people are beginning to generate information about events they observe and participate in. A significant fraction of this information contains multimedia data to share the experience with their audience. A systematic information modeling and management framework is necessary to capture this widely heterogeneous, schemaless, potentially humongous information produced by many different people. This book is an attempt to examine the modeling, storage, querying, and applications of such an event management system in a holistic manner. It uses a semantic-web style graph-based view of events, and shows how this event model, together with its query facility, can be used toward emerging applications like semi-automated storytelling.

How do we go about remembering 7/7? Which aspects of the tragedy – at once an attack on a capital city, on the UK, on ‘the West’– are remembered, and why? What practices and forms are employed as means of commemorating the bombings? And what can all this tell us about the nature of contemporary remembrance? This special issue of Memory Studies seeks to explore these issues by focusing on the mediated commemoration of the 2005 London bombings.

Standard algorithms for template-based in- formation extraction (IE) require predefined template schemas, and often labeled data, to learn to extract their slot fillers (e.g., an embassy is the Target of a Bombing tem- plate). This paper describes an approach to template-based IE that removes this require- ment and performs extraction without know- ing the template structure in advance. Our al- gorithm instead learns the template structure automatically from raw text, inducing tem- plate schemas as sets of linked events (e.g., bombings include detonate, set off, and de- stroy events) associated with semantic roles. We also solve the standard IE task, using the induced syntactic patterns to extract role fillers from specific documents. We evaluate on the MUC-4 terrorism dataset and show that we in- duce template structure very similar to hand- created gold structure, and we extract role fillers with an F1 score of .40, approaching the performance of algorithms that require full knowledge of the templates.

Nested event structures are a common occur- rence in both open domain and domain spe- cific extraction tasks, e.g., a “crime” event can cause a “investigation” event, which can lead to an “arrest” event. However, most cur- rent approaches address event extraction with highly local models that extract each event and argument independently. We propose a simple approach for the extraction of such structures by taking the tree of event-argument relations and using it directly as the representation in a reranking dependency parser. This provides a simple framework that captures global prop- erties of both nested and flat event structures. We explore a rich feature space that models both the events to be parsed and context from the original supporting text. Our approach ob- tains competitive results in the extraction of biomedical events from the BioNLP’09 shared task with a F1 score of 53.5% in development and 48.6% in testing.

In Bloodlands, a brave and original history of mass killing in the twentieth century, Snyder argues that we still lack any real knowledge of what happened in the eastern half of Europe in the twentieth century. And he is right: if we are American, we think “the war” was something that started with Pearl Harbor in 1941 and ended with the atomic bomb in 1945. If we are British, we remember the Blitz of 1940 (and indeed are commemorating it energetically this year) and the liberation of Belsen. If we are French, we remember Vichy and the Resistance. If we are Dutch we think of Anne Frank. Even if we are German we know only a part of the story.

Snyder’s ambition is to persuade the West—and the rest of the world—to see the war in a broader perspective. He does so by disputing popular assumptions about victims, death tolls, and killing methods—of which more in a moment—but above all about dates and geography.

Yes, it's true that the accepted date of 7 September 1940 as the start of the London Blitz is a bit misleading, since there was a non-trivial amount of bombing before that date (e.g. see here). Judging from contemporary press accounts, 7 September certainly seemed to mark an important change in German bombing strategy, but more one of quantity than quality -- almost more an inflection point than a turning point. In retrospect we tend not to see it that way, which is fine. But we could recognise that -- leaving aside the eventual reification involved in the name 'the Blitz' itself -- the 'start of the Blitz' was less clearly defined then than it seems now.
But this is not what the Wikipedia article is talking about. Instead it chooses an equally precise date for the start of the Blitz, 6 September, and says that this is more accurate than 7 September.

Chronos Timeline is designed specifically for needs in the humanities and social sciences to represent time-based data. Chronos allows scholars and students to dynamically present historical data in a flexible online environment. Switching easily between vertical and horizontal orientations, researchers can quickly scan large number of events, highlight and filter events based on subject matter or tags, and recontextualize historical data.

The Story Drifter is a tool that allows a person or a group of people to construct a non-linear narrative based on the context of an event and its connections with surrounding information. Our proposed tool uses stories, photos, videos, historical artifacts, names, dates, and anything else that could help to illustrate not only what happened, but why it happened.

The Living Stories project is an experiment in presenting news, one designed specifically for the online environment. The project was developed by Google in collaboration with two of the country's leading newspapers, The New York Times and The Washington Post.

We try not to stray into neighboring cities, but the video of the protest at Whole Foods Oakland is just too wonderful to miss (hat tip Lisa). It’s a strong, performance-art-like response to Whole Foods CEO John Mackey’s opposition to healthcare reform.

It doesn’t reach the extraordinary brilliance of the Sound of Music performance in Antwerp (below), but I would have liked to have been in Whole Foods when it happened.

Surprisingly, there is still no service that provides standardised URIs for temporal periods. This is partly due to the frequent controversy associated in pinning their boundaries down, as well as the contrast between those with relative and those with absolute dating. CommonEras is an attampt to build a community-based vocabulary service, with appropriate visualisation and search technologies, for temporal periods ('the Sixties', 'the Victorian Era', 'the Golden Age of Comics', etc.).

Survey of approaches to handling time-sensitive statements in RDF.

Assured Date-Period: 5/5/1,789 - 11/9/1,799

We make it easy to explore the events in the timeline, just move your mouse over an event and a pop-up will appear with a longer description and a link to search for related pages. Beyond people, timelines can be a useful tool for displaying information about a period in history, such as the Great Depression. Or a famous sports arena, like Madison Square Garden. Or, say, the highest bridge in the world, the Millau Viaduct.

BarCharts began in 1991, with Bobbie Ford’s handwritten flow chart of Constitutional Law. Today, we produce 400+ QuickStudy laminated quick-reference guides.

"History" subheading lists major events in country's history.

The Timeline Index : People, Periods, Places and Events in a chronological context.

To find out who was there with a given person, I had to first figure out the discrete ranges of the chosen person’s visits. This was not itself easy, because the data was often incomplete (sometimes missing a value for the day, sometimes for the month, sometimes for the year, etc.; sometimes there wouldn’t even be a departure date or arrival date). The dates were a nightmare to work with.

"On the the Wikipedia page for what_happened they gonna show Cornelius leavin' this club with that dancer just now"

Events and periods in the subject index to Irish History Online.

In the First Person is a free, high quality, professionally published, in-depth index of close to 4,000 collections of personal narratives in English from around the world.

The calculus defines possible relations between time intervals and provides a composition table that can be used as a basis for reasoning about temporal descriptions of events.

Web of Fate is a social experiment that harnesses the collective intelligence of the web to visualize and uncover hidden relationships among future and historical events.

Specs for EventsML-G2.

Allows users to interactively browse historical events in a number of ways. It is a generative browser, allowing users to not only view preset collections of events, but to construct their own views of the events based on selected criteria.

Aimed at integrating, combining, and deducing information on the early American history to assist general users or historians in exploring American history by using new technology offered by the Semantic Web.

The case represents a problem-event that has animated some kind of judgment. Any enigma could do-a symptom, a crime, a causal variable, a situation, a stranger, or any irritating obstacle to clarity. What matters is the idiom of the judgment.

Listed are the most important incidents of The Troubles and subsequent peace process.

Human understanding of history, science, culture, and even personal experience, identifies “events” as a central organizing concept.

In the trials, the attorneys disputed the applicable meaning of the term event.

Organized in terms of Presidential administrations by year, with "contemporary events" listed alongside.

A chronicle of medieval Ireland. The entries span the years between AD 431 and AD 1540.

Historian Xia Chun-tao, vice director of one of China’s core ideological think tanks, says, “It’s very natural for historians to have different views on events. But there is only one correct and accurate interpretation..."

Informational graphics that tell stories about subjects, time periods and events.

QVIZ will research and create a framework for visualizing and querying archival resources by a time-space interface based on maps and emergent knowledge structures.

Discourse DB is a new kind of site: a database that uses the power and ease-of-use of wiki technology to collect the opinions of the world's journalists and commentators about ongoing political events and issues.

The Rethinking TimeLines project aims to develop methodologies for modeling and visualising historical events with both temporal (timeline) and spatial (mapping) components.

Dandelife is a place for you to record the events of your life, past and present, public and private.

The research described here aims at building time-varying 3D models that can serve to pull together large collections of images pertaining to the appearance, evolution, and events surrounding one place or artifact over time.

The first 24x7 coverage of a global sporting event using the internet. The media generated by the boats in their race around the world—emails, photos, videos, position data—was used for data-driven storytelling.

Querying by location, time and/or pollutant concentration is useful for researchers and others familiar with air quality data. But for people not familiar with what PM2.5 is or the significance of 65 ug/m3, searching by event type is useful.

I-School professor whose research includes event gazetteers, narrative structures, and news archives.

Most news websites could let you see the stories they published on any particular day but few seem to offer such a view. Browsing events in this way is either deemed unpopular or not something sites want you to do.

It The BBC is well placed to try and weave the explosion of personal content into a comprehensive narrative that mixes the best of the BBC's archived output with the best of the collective memory and 'citizen history' that they can tap into.

Users can browse hundreds of key events & discover how the past has shaped the world we live in today.

The MBA RSS Edited Feeds project is intended to create feeds from member blogs by subject, geography, both - or events.

Discussion of how to formally represent the significance of news events to specific entities, for facilitating the machine processing of news.