scholarly journals United we fall: All-or-none forgetting of complex episodic events

Author(s):  
Bardur Hofgaard Joensen ◽  
M. Gareth Gaskell ◽  
Aidan J Horner

Do complex event representations fragment over time or are they, instead, forgotten in an all-or-none manner? For example, if we met a friend in a café and they gave us a present, do we forget the constituent elements of this event (location, person and object) independently, or would the whole event be forgotten as one? Research suggests that item-based memories are forgotten in a fragmented manner. However, we do not know how more complex episodic, ‘event-based’ memories are forgotten. We assessed both retrieval accuracy and dependency – the statistical association between the retrieval successes of different elements from the same event – for complex events. Across 4 experiments, we show that retrieval dependency is found both immediately after learning and following a 12-hour and 1-week delay. Further, the amount of retrieval dependency after a delay is greater than that predicted by a model of independent forgetting. This dependency was only seen for coherent ‘closed-loops’, where all pairwise associations between locations, people and objects were encoded. When ‘open-loops’ were learnt, where only two out of the three possible associations were encoded, no dependency was seen immediately after learning nor after a delay. Finally, we also provide evidence for higher retention rates for closed-loops than open-loops. Therefore, closed-loops do not fragment as a function of forgetting, and are retained for longer than open-loops. Our findings suggest that coherent episodic events are not only retrieved, but also forgotten, in an all-or-none manner.


2016 ◽  
Vol 17 (1) ◽  
pp. 48-76 ◽  
Author(s):  
Anne-Laure Mealier ◽  
Grégoire Pointeau ◽  
Peter Gärdenfors ◽  
Peter Ford Dominey

Abstract In robotics research with language-based interaction, simplifications are made, such that a given event can be described in a unique manner, where there is a direct mapping between event representations and sentences that can describe these events. However, common experience tells us that the same physical event can be described in multiple ways, depending on the perspective of the speaker. The current research develops methods for representing events from multiple perspectives, and for choosing the perspective that will be used for generating a linguistic construal, based on attentional processes in the system. The multiple perspectives are based on the principle that events can be considered in terms of the force driving the event, and the result obtained from the event, based on the theory of Gärdenfors. In addition, within these perspectives a further refinement can be made with respect to the agent, object, and recipient perspectives. We develop a system for generating appropriate construals of meaning, and demonstrate how this can be used in a realistic dialogic interaction between a behaving robot and a human interlocutor.



in which members share little in common perceptually. Food consists simply of those items that play a certain role in children's breakfast, lunch, and dinner scripts. In an especially well-known study, Lucariello, Kyratzis, and Nelson (1992) asked preschool children of various ages to provide specific items for five super-ordinate categories: food, clothes, animals, furniture, and tools. The first three of these in particular were hypothesized to have slot-filler structure because of their participation in salient events in children's lives, and indeed, it was found that the basis for each of these categories for young children was the similar events in which its exemplars participated. There was also evidence that the older children formed these categories on the basis of more different types of events than younger children. Subsequent research has shown that children can form both syntagmatic and paradigmatic categories from their initial event representations (see Nelson, 1996, for a review). Nelson is one of the only theorists of children's language development who has gone onto focus on the nature of children's lexical development later in the preschool period (the one major exception being Anglin, 1977,1983). Briefly, the idea is that by establishing lexical fields of similar terms, children construct relations such as synonymy, antonymy, and hy pony my (hierarchical relations). The establishment of these relations makes possible "the manipulation of language terms without refer-ence to situational context" (Nelson, 1985, p. 214); that is, children establish lexical relations among words, "unencumbered by all of the syntagmatic entailments of the conceptual system" (Nelson, 1985, p. 214). Establishing these kinds of abstract rela-tions enables children to, among other things, perform in adult-like ways in explicit verbal classification tasks as they approach school age. It is only at this point that Nelson is willing to say that children have "a system of semantic relations that is purely symbolic and semiautonomous, that is, it can operate independently of the conceptual system" (Nelson, 1985, p. 214). Strong evidence for this proposal was re-cently supplied by Sell (1992). In a study of children ranging in age from 2 to 10 years, she found that the youngest children seemed to possess mainly categories based in specific events. The slightly older children (5-6 years of age) possessed, in addition, slot-filler categories based on participant roles in whole classes of events. It was only the oldest, school-aged children, who possessed fully taxonomic concep-tual categories independent of specific events and event types. With respect to the grammatical structure of language, Tomasello (1992a) used Nelson's event-based model to explicate some aspects of children's early multi-word productions. The hypothesis was that the basic structure of children's earliest multiword utterances is provided by verbs. The defining feature of verbs is of course the dynamic and sequential nature of their underlying conceptualizations; they refer to events and states of affairs. Moreover, the meaning of a verb perforce includes participant roles such as agent and patient as an integral component. For example, the meaning of the verb give includes the giver, the thing given, and the person given to as they engage in certain activities. Children's understanding and



Author(s):  
Ajay Kaushik ◽  
Ravi Teja Yakkali ◽  
S. Indu ◽  
Feroz Ahmed ◽  
Daya Gupta ◽  
...  

Sensing and data aggregation capabilities of wireless sensor networks (WSNs) depends on efficient deployment of sensor nodes (SNs) in an area. In a large surveillance space, there is a need for more SNs to cover important crucial events despite of the optimum coverage. The authors propose an event-based efficient deployment algorithm (EEDA) for relocation of redundant sensors to the event location to achieve full coverage. They divide the deployment region into small square cells that allows individual cells to be efficiently monitored, instead of considering the whole scenario as one unit. EEDA ensures efficient coverage of the entire deployment region and senses the occurrence of any static or dynamic event with an optimum number of sensors. EEDA with square cells performs better than existing hexagon cell algorithm by 39%. EEDA is validated by simulation as well as by experimental results.



2020 ◽  
Author(s):  
Jianqiu Zhang

In this paper, the human conceptual system is treated as a dynamic system consisting of various functional modules. It is delineated that episodic events are input units to the system, which explains why concept meanings are contextual. Then it is elucidated that innate analyzers outputs form primitive conceptual nodes through an abstraction and segmentation process. Furthermore, it is recognized that primitive conceptual nodes act as indices for retrieving relevant events to enable event-based reasoning system (CS0) in a three-tiered structure in addition to the widely known human conceptual system 1 (CS1) and system 2 (CS2). This three-tiered theory addresses the important question of how it is possible to conduct reasoning before concept prototypes and their associative links are formed. Major functional modules of CS0, CS1, and CS2 have been identified, particularly, there exists a Dynamic Event Composition Module (DECM) that takes activated conceptual nodes as inputs and outputs composed events as the basis for CS1 reasoning. This account explains the CS1 characteristics documented before. Furthermore, data types such as knowledge systems, variables, pointers, and linked lists, are identified for CS2 as the basic constructs that support logical reasoning. Various descriptive concept contents can be tied into the system structure neatly: exemplars are embedded in the episodic events at the CS0 level; concept prototypes and their associative links are generated at the CS1 level. Theories of concepts are stored in the CS2 level for logical reasoning. The discussion has been organized from the conceptual system structure perspective, the conceptual node structure perspective, the reasoning perspective, the evolutionary and developmental perspective, the concept composition perspective, and the subjective and objective perspective. Through this multi-tiered approach, many conflicting descriptions of the human conceptual system have been resolved by attributing them to their appropriate tiers.



Author(s):  
A. S. Delsouc ◽  
M. E. Barber ◽  
W. Perez-Martinez ◽  
I. Briceño-De-Urbaneja

Abstract. Restricted episodic changes occurred in a short time period and over large spatial extents. Extreme weather conditions usually give rise to restricted episodic changes. Sentinel-1 radar images of the Salar de Aguas Calientes in Chile acquired in the Altiplanic winter (March 2015 and June 2017), 2018 austral winter and 2017–2018 springtime, demonstrates the ability to monitor episodic events remotely. The results of the backscattered power are encouraging and show episodic variations in VV polarization at C-band. The surface features in Salar de Aguas Calientes Sur change in response to snowfalls during either the Altiplanic or austral winter with an increase of the backscattering in presence of dry snow over the salt pan crust. Flooding events related to snow-melting during spring 2017–2018 showed a decrease in the backscattering signal over ponded water and an increase when wind blows over the water. Remote-sensing observations of the salar can provide a means for monitoring changes in the surface of the salar and a better understanding of the associated climatic episodic event processes. Furthermore, it can help to have a better understanding of environmental changes in arid regions and the understanding of global climate change.



1992 ◽  
Vol 72 (4) ◽  
pp. 441-452 ◽  
Author(s):  
J. Windsor ◽  
T. R. Moore ◽  
N. T. Roulet

Measurements of methane flux from northern peatlands to the atmosphere are complicated by high spatial and temporal variability. We quantified the variability of methane flux from two subarctic fens near Schefferville, northern Quebec, using a static chamber technique. Within the seasonal pattern of increasing fluxes associated with the warming of the peat profile, episodic fluxes of methane were observed. One set of episodic events occurred during the spring thaw of the upper layers of the peat, which released methane stored in ice over the winter. The second set, generally of short duration (< 2 d), occurred in mid-summer and appeared to be related primarily to the lowering of the water table. In four of six subjectively-identified, episodic fluxes during the 1990 summer the flux during the episodic event was equal to or greater than the upper 95% confidence level of the three fluxes before and after the event (t-statistic probabilities ranged from < 0.001 to 0.038). Mechanisms to account for these episodic fluxes of methane include increased methane diffusivity, removal of overburden pressure and reduced rates of methane consumption in the surface layers of the peat. Omission of these episodic fluxes could lower estimates of seasonal methane emissions by 7–22%. Key words: Peatlands, fens, methane



Author(s):  
G.D. Danilatos

The advent of the environmental SEM (ESEM) has made possible the examination of uncoated and untreated specimen surfaces in the presence of a gaseous or liquid environment. However, the question arises as to what degree the examined surface remains unaffected by the action of the electron beam. It is reasonable to assume that the beam invariably affects all specimens but the type and degree of effect may be totally unimportant for one class of applications and totally unacceptable for another; yet, for a third class, it is imperative to know how our observations are modified by the presence of the beam. The aim of this report is to create an awareness of the need to initiate research work in various fields in order to determine the guiding rules of the limitations (or even advantages) due to irradiation.



Author(s):  
G.F. Bastin ◽  
H.J.M. Heijligers ◽  
J.M. Dijkstra

For the calculation of X-ray intensities emitted by elements present in multi-layer systems it is vital to have an accurate knowledge of the x-ray ionization vs. mass-depth (ϕ(ρz)) curves as a function of accelerating voltage and atomic number of films and substrate. Once this knowledge is available the way is open to the analysis of thin films in which both the thicknesses as well as the compositions can usually be determined simultaneously.Our bulk matrix correction “PROZA” with its proven excellent performance for a wide variety of applications (e.g., ultra-light element analysis, extremes in accelerating voltage) has been used as the basis for the development of the software package discussed here. The PROZA program is based on our own modifications of the surface-centred Gaussian ϕ(ρz) model, originally introduced by Packwood and Brown. For its extension towards thin film applications it is required to know how the 4 Gaussian parameters α, β, γ and ϕ(o) for each element in each of the films are affected by the film thickness and the presence of other layers and the substrate.



Sign in / Sign up

Export Citation Format

Share Document