Figure 4 - uploaded by Paulo Oliveira
Content may be subject to copyright.
Source publication
Abstract Information Summarization is one of the key challenges for current and future information systems. In this paper, we will outline a system that comprises modules for summarizing texts and time series to study the link between the two. Summaries of texts are generated using a lexical analysis of cohesion in texts focusing on key sentences t...
Contexts in source publication
Context 1
... textual dataset comprises six financial news stories per day in May 2002 whereas the numeric data is the FTSE 100 index for the same period. Figure 4 below shows the FTSE 100 time series for May 2002 with the trend (green line) and three most significant turning points (circled red) identified by our system. Table 1 shows our system's summary output (verbal signal generation) for other features extracted from the FTSE 100 time series data for May 2002. ...
Context 2
... 1 shows our system's summary output (verbal signal generation) for other features extracted from the FTSE 100 time series data for May 2002. By looking at the system output for the FTSE 100 ( Figure 4 and Table 1) we conclude that May 15, 2002 signifies a major turning point in that the trend is an uptrend until this point and a downtrend thereafter. ...
Context 3
... Table 2, May 14 summary is good in the sense that it talks about the FTSE 100 "opening higher" the next day. This is supported well by the FTSE 100 index, which shows a peak the next day (May 15) in Figure 4. However, the time series module defines May 15 as a turning point and hence the index or the market is expected to fall from this point onwards. ...
Similar publications
Methods for detecting and quantifying nonlinearities in nonstationary time
series are introduced and developed. In particular, higher-order wavelet
analysis was applied to an ideal time series and the quasi-biennial oscillation (QBO)
time series. Multiple-testing problems inherent in wavelet
analysis were addressed by controlling the false discover...
In order to get insight into the spatial structure of turbulent fluctuations on an inner flux surface of a toroidal plasma, measurements with a poloidal probe array were carried out in the torsatron TJ-K. A local maximum in fluctuation levels and net transport was found on the LFS of the magnetic configuration, where bad curvature gives rise to enh...
Summer-spring predominance of tuberculosis (TB) has been widely reported. The relative contributions of exogenous recent infection versus endogenous reactivation to such seasonality remains poorly understood. Monthly TB notifications data between 2005 and 2017 in Hong Kong involving 64,386 cases (41% aged ≥ 65; male-to-female ratio 1.74:1) were exa...
One major candidate of lithosphere-atmosphere-ionosphere (LAI) coupling mechanism is through atmospheric oscillations triggered near Earth’s surface due to some pre-earthquake (EQ) effect, and this channel has been extensively proved by using meteorological disturbances much more easily treated than pre-EQ effects (Korepanov et al., 2009). In the p...
Citations
... The increasing use of multimedia also raises a need for multimodal approaches that are able to extract important content from images or videos and include them into a summary. For multimodal summarization, there are yet only few case studies for a few domains, such as financial reports (Ahmad et al. 2004). Among the biggest challenges is, however, the heterogeneity of the individual postings, which makes the task of live blog summarization much different to multi-document summarization of multiple news articles covering very similar information or microblog summarization of a large number of highly redundant posts. ...
Live blogs are an increasingly popular news format to cover breaking news and live events in online journalism. Online news websites around the world are using this medium to give their readers a minute by minute update on an event. Good summaries enhance the value of the live blogs for a reader, but are often not available. In this article, (a) we first define the task of summarizing a live blog, (b) study ways of automatically collecting corpora for live blog summarization, and (c) understand the complexity of the task by empirically evaluating well-known state-of-the-art unsupervised and supervised summarization systems on our new corpus. We show that live blog summarization poses new challenges in the field of news summarization, since frequency and positional signals cannot be used. We make our tools publicly available to reconstruct the corpus and to conduct our empirical experiments. This encourages the research community to build upon and replicate our results.
This paper proposes a system for retrieving time-series data based on a linguistic query given by a user. Our proposed system uses a line chart as a query. The system generates a linguistic query by verbalizing the chart first, and then it retrieves similar charts by using the obtained linguistic query.
It is more important than ever to be able to extract relevant information and to understand or review its content quickly.
Spotting trends is one such way of understanding or reviewing. We propose multi-modal summarization for trend information
as a new technology to help this activity. We also describe the design of a data set that will foster research in the field,
and introduce an ongoing unique workshop using this data set.
Time series is an important class of temporal data objects and it can be easily obtained from scientific and financial applications. A time series is a collection of observations made chronologically. The nature of time series data includes: large in data size, high dimensionality and necessary to update continuously. Moreover time series data, which is characterized by its numerical and continuous nature, is always considered as a whole instead of individual numerical field. The increasing use of time series data has initiated a great deal of research and development attempts in the field of data mining. The abundant research on time series data mining in the last decade could hamper the entry of interested researchers, due to its complexity. In this paper, a comprehensive revision on the existing time series data mining research is given. They are generally categorized into representation and indexing, similarity measure, segmentation, visualization and mining. Moreover state-of-the-art research issues are also highlighted. The primary objective of this paper is to serve as a glossary for interested researchers to have an overall picture on the current time series data mining development and identify their potential research direction to further investigation.
Many working processes are complex and composed by heterogeneous atomic tasks, e.g. editing, assembling data from different sources (as databases or laboratory's devices) with texts, images or learning objects, or submitting them to processing components to retrieve information, to render them, re-format, submit to computations, and other types of information processing. All these complex processes are extremely difficult to be modeled and automatized without having a flexible, multimodular evolutionary system in place. Support to information from different modalities increases the performance of a computer system originally designed for a task with a unimodular nature. In this paper, we discuss the idea of task management system (TMS) as a component based system which offers a virtual workbench to search, acquire, describe and assemble computational agents performing single autonomous tasks into working processes. We sustain that TMS is a cutting edge platform to develop SW solutions for problems related to workflow automatization and design. The architecture we propose follows the conceptual track of the TMS to allow composition and arrangement of atomic modules into a complex system: a core configuration of the system can be extended with a set of task/components, chunks of activities which are considered basic to working flow composition. The workflow designer selects the relevant chunks from system repositories, drags them into a working system area and assembles them into a working process. Complex activities could be formally described, implemented and applied with a consequent advantage for personnel re-organization toward more conceptual activities.
Many working processes are complex and composed by heterogeneous atomic tasks, e.g. editing, assembling data from different sources (as databases or laboratory's devices) with texts, images or learning objects, or submitting them to software components to retrieve information, to render them, re-format, submit to computations, and other types of information processing. All these processes heavily require procedural knowledge which is tacit as owned by experts of the working activity; they are complex and are extremely difficult to be modeled and automatized without having a flexible, multimodular evolutionary system in place. Support to information from different modalities increases the performance of a computer system originally designed for a task with a unimodular nature. In this paper, we discuss the idea of task management system (TMS) as a component-based system which offers a virtual workbench to search, acquire, describe and assemble computational agents performing single autonomous tasks into working processes. We sustain that TMS is a cutting edge platform to develop software solutions for problems related to workflow automatization and design. The architecture we propose follows the conceptual track of the TMS to allow composition and arrangement of atomic modules into a complex system. A configuration of the workflow can be implemented and extended with a set of task/components, chunks of activities which are considered basic elements of the workflow. By interacting with the TMS in editing mode, the workflow designer selects the relevant chunks from system repositories, drags them into a working system area and assembles them into a working process. As the main actor of the system, the workflow designer is provided with an environment resembling an artisan’s workshop, to let her/him select the relevant chunks from system repositories, drags them into a working area and assembles them into a working TMS instance, which represents the working process. Global interaction modality of the TMS instance is moulded and specialized on the base of the specific modalities of the task/components which have been retrieved from the system repositories and each time negotiated. Complex activities could be formally described, implemented and applied with a consequent advantage for personnel re-organization toward more conceptual activities.
