Temporal Recency Research Articles

Recency-based spatio-temporal similarity exploration for POI recommendation in location-based social networks

Point-of-interest (POI) recommendation is one of the primary tasks of location-based social networks (LBSNs). With user data in bulk, extracting useful information and addressing issues such as data sparsity and cold-start problems looming large in collaborative filtering become difficult. One of the plausible solutions is to incorporate contextual information into the recommendation process. In this article, we propose a Recency-based Spatio-Temporal Similarity Exploration (RSTSE) for POI recommendation that utilizes the recency-based trust estimation among the prospective neighbors of the target user. The trust level is categorized into two heads: direct trust, which can be extracted from the peer group information of the user, and indirect trust, which is measured based on venue popularity, temporal recency, radial proximity, and transitivity. The approach consists of two phases. In the incipient phase, POIs are extracted based on the preferences of potential neighbors, including the users who are recognized peers, the users with similar visiting histories in the spatial and temporal context, and the users with friend-of-friend relations. The telic phase involves Neural Collaborative Filtering (NCF) to capture the linear and non-linear user–POI interactions better. RSTSE has been evaluated on three real-world datasets, namely, Gowalla, Foursquare, and Weeplaces, and the results suggest efficacy over other state-of-the-art approaches.

Geolocation Prediction in Twitter Using Social Networks: A Critical Analysis and Review of Current Practice

Geolocated social media data provides a powerful source of information about place and regional human behavior. Because little social media data is geolocation-annotated, inference techniques serve an essential role for increasing the volume of annotated data. One major class of inference approaches has relied on the social network of Twitter, where the locations of a user's friends serve as evidence for that user's location. While many such inference techniques have been recently proposed, we actually know little about their relative performance, with the amount of ground truth data varying between 5% and 100% of the network, the size of the social network varying by four orders of magnitude, and little standardization in evaluation metrics. We conduct a systematic comparative analysis of nine state-of-the-art network-based methods for performing geolocation inference at the global scale, controlling for the source of ground truth data, dataset size, and temporal recency in test data. Furthermore, we identify a comprehensive set of evaluation metrics that clarify performance differences. Our analysis identifies a large performance disparity between that reported in the literature and that seen in real-world conditions. To aid reproducibility and future comparison, all implementations have been released in an open source geoinference package.

Dynamic Value: Temporal Distance and Group Size of Information in Shaping Social Influence

Individuals often pay attention to recency and source size of reference information when learning from it to aid own decisions. Despite research on the independent effects of temporal distance and group size on social influence of information, we have limited understanding about how recency and group size interact with each other to shape the influence of temporally dispersed information on individual decision making. Using 6.3 million single family listing records in California from 2009 to 2018, I find that in addition to their independent positive effects on social influence, temporal recency and group size are also contingent on each other. The discrepancy in the relative importance of recent information and distant information diminishes as the group size increases. Decision makers focus more on general patterns instead of individual recent information with the extension of time windows. Intercorrelation of information enhanced by temporal structure in large time windows also facilitates selection behaviors when incorporating information. Collectively, social influence forms a U-shape across time windows. These findings highlight the importance of accounting for the temporal structure of information for social influence research and provide new insights into the social dynamics of decision making process.

“Past injustice and present prejudice”: Reducing racial bias and increasing sympathy by framing historical racism as recent

Racial privity judgments – or the perceived causal connection between historical racial discrimination and current suffering among Black Americans – predicts sympathy for the victims of past injustices and perceptions of contemporary racial inequality. Four studies investigated the ideological roots of privity judgments; focusing on subjective temporal perceptions associated with privity judgments (e.g., subjective perceptions that past discrimination occurred more, versus less, recently). Study 1 revealed that liberals perceived historical instances of racial discrimination as having occurred more recently than conservatives, and that temporal perceptions of recency were associated with less anti-Black bias. Studies 2–4 manipulated temporal perceptions of recency by framing past discrimination as having occurred more recently. Results revealed that increasing perceived temporal recency resulted in reduced anti-Black bias and greater sympathy for present-day victims of racial discrimination across political ideology. Discussion surrounds how framing historical information as subjectively recent has implications for prejudice reduction.

Forecasting Spatio-Temporal Variation in Residential Burglary with the Integrated Laplace Approximation Framework: Effects of Crime Generators, Street Networks, and Prior Crimes

ObjectivesWe investigate the spatio-temporal variation of monthly residential burglary frequencies across neighborhoods as a function of crime generators, street network features and temporally and spatially lagged burglary frequencies. In addition, we evaluate the performance of the model as a forecasting tool.MethodsWe analyze 48 months of police-recorded residential burglaries across 20 neighborhoods in Amsterdam, the Netherlands, in combination with data on the locations of urban facilities (crime generators), frequencies of other crime types, and street network data. We apply the Integrated Laplace Approximation method, a Bayesian forecasting framework that is less computationally demanding than prior frameworks.ResultsThe local number of retail stores, the number of street robberies perpetrated and the closeness of the local street network are positively related to residential burglary. Inclusion of a general spatio-temporal interaction component significantly improves forecasting performance, but inclusion of spatial proximity or temporal recency components does not.DiscussionOur findings on crime generators and street network characteristics support evidence in the literature on environmental correlates of burglary. The significance of spatio-temporal interaction indicates that residential burglary is spatio-temporally concentrated. Our finding that recency and proximity of prior burglaries do not contribute to the performance of the forecast, probably indicates that relevant spatio-temporal interaction is limited to fine-grained spatial and temporal units of analysis, such as days and street blocks.

Remembering faces: The effects of emotional valence and temporal recency

It is known that the longer an information has been memorized, the stronger is its memory trace, and that emotionally-valenced information is more solid than neutral one. We investigated whether the emotional content of recent information might enhance its memory, making it as familiar as information known for a long time. We compared ERPs alternately recorded in response to old and solid information from long term memory (i.e., faces of popular movie stars), to recently acquired emotional information (faces of fictional characters), and to completely new information (faces of previously unknown people). Initially participants familiarized with the fictional police dossiers of 10 victims of dramatic deaths (recent faces), twice a day for seven days before EEG recordings. Recent faces were compared with faces of movie stars and unknown faces in an old/new recognition task. N200 and FN400 responses were affected by face familiarity (with no difference between old and recent faces), while parietal late positivity (LP) was sensitive to temporal recency, being it greater to old than recent faces. Interestingly, LP amplitude was similar for old and recent own-sex faces (victims) that were therefore equally memorable. It is shown that emotional memory can overcome temporal recency thus improving memory recall.

Perceptual Quality of HTTP Adaptive Streaming Strategies: Cross-Experimental Analysis of Multi-Laboratory and Crowdsourced Subjective Studies

Today’s packet-switched networks are subject to bandwidth fluctuations that cause degradation of the user experience of multimedia services. In order to cope with this problem, HTTP adaptive streaming (HAS) has been proposed in recent years as a video delivery solution for the future Internet and being adopted by an increasing number of streaming services, such as Netflix and Youtube. HAS enables service providers to improve users’ quality of experience (QoE) and network resource utilization by adapting the quality of the video stream to the current network conditions. However, the resulting time-varying video quality caused by adaptation introduces a new type of impairment and thus novel QoE research challenges. Despite various recent attempts to investigate these challenges, many fundamental questions regarding HAS perceptual performance are still open. In this paper, the QoE impact of different technical adaptation parameters, including chunk length, switching amplitude, switching frequency, and temporal recency, are investigated. In addition, the influence of content on perceptual quality of these parameters is analyzed. To this end, a large number of adaptation scenarios have been subjectively evaluated in four laboratory experiments and one crowdsourcing study. A statistical analysis of the combined data set reveals results that partly contradict widely held assumptions and provide novel insights in perceptual quality of adapted video sequences, e.g., interaction effects between quality switching direction (up/down) and switching strategy (smooth/abrupt). The large variety of experimental configurations across different studies ensures the consistency and external validity of the presented results that can be utilized for enhancing the perceptual performance of adaptive streaming services.

Context-specific memory in children with ADHD

Objectives:Memory for where and when information is learned, or contextual aspects of memory, as opposed to fact or item memory, has been suggested to be more dependent on frontal brain and related executive control systems. Children with neurodevelopmental disorders involving frontal brain systems have been noted to have problems on contextual memory tasks. Children with attention deficit hyperactivity disorder (ADHD) have difficulties with forgetfulness and are known to have deficits in aspects of executive control with associated dysfunction in frontal brain systems; therefore, while their item memory is not typically impaired, they may have difficulty on tasks requiring contextual memory abilities. The goal of this study was to compare children with ADHD to typically developing children, on two different contextual memory tasks (source memory and memory for temporal recency) and to compare this with their respective performance on traditional fact/item memory tasks.Methods:Participants included 36 children with combined type ADHD and 36 typically developing controls ages 6–13 years, matched by age and gender. In addition to a measure of intellectual ability, children completed two different contextual memory tasks; one assessing recall for the source of new information and one assessing recall of the temporal context of information (recency). Incorporated into these tasks were standard measures of item/fact memory recall.Results:Results revealed that children with ADHD performed similarly to control participants on memory recognition tasks, but differed from controls on the contextual memory tasks. Children with ADHD outperformed control participants when recalling the source from which they had learned information, but performed more poorly than control on a task that required them to make a recency/temporal order judgment.Conclusion:The data suggest that while children with ADHD may have somewhat similar capacity to recognise new information, they differ in aspects of their performance on contextual aspects of memory. These differences are not uniform and mirror neuroscience findings supporting dissociable contextual memory systems. Potential theoretical explanations for the dissociations seen are discussed.

Becoming a better person: temporal remoteness biases autobiographical memories for moral events.

Our autobiographical self depends on the differential recollection of our personal past, notably including memories of morally laden events. Whereas both emotion and temporal recency are well known to influence memory, very little is known about how we remember moral events, and in particular about the distribution in time of memories for events that were blameworthy or praiseworthy. To investigate this issue in detail, we collected a novel database of 758 confidential, autobiographical narratives for personal moral events from 100 well-characterized healthy adults. Negatively valenced moral memories were significantly more remote than positively valenced memories, both as measured by the valence of the cue word that evoked the memory as well as by the content of the memory itself. The effect was independent of chronological age, ethnicity, gender or personality, arguing for a general emotional bias in how we construct our moral autobiography.

Inclusion-exclusion of positive and negative past selves: Mood congruence as information.

The current research challenges the widespread truism that recalling a positive self necessarily increases self-esteem, whereas recalling a negative self necessarily decreases self-esteem. Four experiments demonstrate that chronically happy people show a relative increase in self-esteem by recalling either a positive or a negative self. Chronically sad people, however, show a relative decrease in self-esteem by recalling either a positive or a negative self. These effects are due to divergent perceptions of mood congruence between the recalled self and the current self. Specifically, happy people perceive high mood congruence between a recalled positive self and the current self but low mood congruence between a recalled negative self and the current self. In contrast, sad people perceive high mood congruence between a recalled negative self and the current self but low mood congruence between a recalled positive self and the current self. Independent of chronic mood, mood congruence leads to perceptions of temporal recency, whereas mood incongruence leads to perceptions of temporal distance. In line with the inclusion-exclusion model of social judgment, perceived temporal recency elicits assimilation effects on self-esteem, whereas perceived temporal distance elicits contrast effects on self-esteem.

Goal-dependent modulation of declarative memory: Neural correlates of temporal recency decisions and novelty detection

Declarative memory allows an organism to discriminate between previously encountered and novel items, and to place past encounters in time. Numerous imaging studies have investigated the neural processes supporting item recognition, whereas few have examined retrieval of temporal information. In the present study, functional magnetic resonance imaging (fMRI) was conducted while subjects engaged in temporal recency and item novelty decisions. Subjects encountered three-alternative forced-choice retrieval trials, each consisting of two words from a preceding study phase and one novel word, and were instructed to either identify the novel item (Novelty trials) or the more recently presented study item (Recency trials). Relative to correct Novelty decisions, correct Recency decisions elicited greater activation in a network of left-lateralized regions, including frontopolar and dorsolateral prefrontal cortex and intraparietal sulcus. A conjunction analysis revealed that these left-lateralized regions overlapped with those previously observed to be engaged during source recollection versus novelty detection, suggesting that during Recency trials subjects attempted to recollect event details. Consistent with this interpretation, correct Recency decisions activated posterior hippocampus and parahippocampal cortex, whereas incorrect Recency decisions elicited greater anterior cingulate activation. The magnitude of this latter effect positively correlated with activation in right dorsolateral prefrontal cortex. Finally, correct Novelty decisions activated the anterior medial temporal lobe to a greater extent than did correct Recency decisions, suggesting that medial temporal novelty responses are not obligatory but rather can be modulated by the goal-directed allocation of attention. Collectively, these findings advance understanding of how subjects strategically engage frontal and parietal mechanisms in the service of attempting to remember the temporal order of events, and how retrieval goals impact novelty processing within the medial temporal lobe.

Planning causes and consequences in discourse

We investigate the planning of cause and consequence in language production by examining participants’ continuations to discourse fragments in four experiments. Our studies indicate how the content of the continuation, and the association between the continuation and prior text, are influenced by the nature of prior discourse. People tend to anchor upcoming utterances to information that is both temporally recent and textually recent (Experiment 1). Furthermore, when discourse fragments describe the cause of an event, continuations tend to describe its consequence, and vice versa (Experiments 2 and 3). Finally, people plan utterances based on prior causality information drawn from world knowledge about the typicality of events (Experiment 4). When this knowledge provides information about causes, people are more likely to plan an upcoming consequence. Hence, people seek to satisfy gaps in their discourse model of the unfolding narrative, and use features of textual and temporal recency to anchor their productions.

Launching the effect: representations of causal movements are influenced by what they lead to.

We investigated whether the representation of an observed causal movement is influenced by its observed effect. Subjects watched displays showing collisions between two objects. In this "launching event" (Michotte, 1946/1963), one of the two objects (Object A) started to move and set a second, initially stationary, object (Object B) into motion, which gave a strong impression of apparent causality. The apparent effectiveness of A's movement was manipulated by varying the velocities of A and B. When the velocity of B was higher than that of A, the effectiveness of the collision was high; when it was smaller it was low. Then, subjects were asked to reproduce the velocity of the causal movement. Reproduced velocity followed the velocity of both Object A and Object B, which supports the hypothesis that the effect of a movement is integrated with its apparent cause. However, when apparent causality was reduced by changing the direction of motion of B or by covering the point of collision, the influence of the effect on the representation of the cause persisted, suggesting that retroactive interference may account for the findings. The interference effect could not be reduced to temporal recency or spatial integration and was not obtained in the reverse temporal order (proactive interference). Rather, the two successive movements were blended in memory.

Use of temporal context information in Azheimer's disease

Current theories of memory and memory dysfunction emphasize the importance of the use of contextual information in the storage and retrieval of specific episodes. Patients with lesions primarily in the diencephalon perform poorly on a measure of temporal recency, while those with temporal lobe amnesia are significantly better. The present study evaluated the ability of Probable AD patients to utilize information about temporal recency in the context of a recognition memory task. The performance of the AD patients was significantly impaired and resembled that of patients with Korsakoff's syndrome. The pattern of performance suggested that they were able to recognize which stimuli had been previously presented, but were unable to recall the correct temporal context. Further, there were no significant correlations between measures of temporal recency and those of frontal lobe function. These data suggest that AD patients, like other groups of patients with severe memory defects, are impaired at the use of context information in their attempts to remember recent events.

Memory for temporal order and frequency of occurrence in detoxified alcoholics

The ability to retain temporal order (recency) and event frequency information was assessed in long-term alcoholics and in nonalcoholic control subjects. On the temporal recency task, the alcoholics were impaired when they were asked unexpectedly to judge how recently these stimuli had been presented. When subjects were aware that they would eventually be asked to make decisions about temporal order, the alcoholics evidenced significant difficulties only for the figural stimuli. On the task requiring judgements about the frequency of events, alcoholics demonstrated significant deficits only when they had to make unexpected judgements about figural stimuli. The implications of these findings for the neurological basis of alcoholics' cognitive deficits are discussed.

Serial Position Effects in Young Children: Temporal or Spatial?

3 experiments examined the notion that children's primacy and recency in the probe-type pictorial memory task are related to the spatial distinctiveness of the leftmost and rightmost items presented. The correspondence between spatial location and temporal position of pictorial items was varied systematically in each experiment. 3-year old children were observed in experiment 1; 8-year olds in experiment 2; and 2 age groups (5 years and 8 years) in experiment 3. Taken together, results of the 3 studies indicated strong primacy and recency for all age groups when temporal order was completely confounded with left-to-right spatial arrangement of the items. When spatial and temporal components were separated, so that each could be assessed independently of the other, the data showed: (a) consistent temporal recency, (b) weak or absent temporal primacy in all but 1 experiment (experiment 3), (c) little evidence of primacy and recency related solely to spatial location, and (d) no significant interactions of age with serial position (spatial location or temporal position). These data do not support the inference that spatial distinctiveness or anchoring alone is responsible for children's primacy and recency in the widely used probe-type pictorial memory task. Rather, they suggest that preschoolers and older children alike are able to code temporal order along with the corresponding pictorial and location information in this kind of task.