An Indexing Framework for Efficient Visual Exploratory Subgraph Search in Graph Databases

Exploratory search is a search paradigm that plays a vital role in databases, data mining, and information retrieval to assist users to get familiar with the underlying databases. It supports iterative query formulation to explore the data space. Despite its growing importance, exploratory search on graph-structured data has not received adequate attention in the literature. In this paper, we demonstrate a novel system called Vincent that facilitates an efficient exploratory subgraph search in a graph database containing a large collection of small or medium-sized graphs. By automatically generating the content for panels in GUI and diversified patterns from databases and providing a visual result explorer, Vincent supports data-driven visual query formulation, incremental subgraph processing, and efficient query result summarization.

Despite the growing importance of exploratory search, information retrieval (IR) systems tend to focus on lookup search. Lookup searches are well served by optimising the precision and recall of search results, however, for exploratory search this may be counterproductive if users are unable to formulate an appropriate search query. We present a system called PULP that supports exploratory search for scientific literature, though the system can be easily adapted to other types of literature. PULP uses reinforcement learning (RL) to avert the user from context traps resulting from poorly chosen search queries, trading off between exploration (presenting the user with diverse topics) and exploitation (moving towards more specific topics). Where other RL-based systems suffer from the cold start problem, requiring sufficient time to adjust to a user's information needs, PULP initially presents the user with an overview of the dataset using temporal topic models. Topic models are displayed in an interactive alluvial diagram, where topics are shown as ribbons that change thickness with a given topics relative prevalence over time. Interactive, exploratory search sessions can be initiated by selecting topics as a starting point.

Search activities involving knowledge acquisition, investigation and synthesis are collectively known as exploratory search. Exploratory search is challenging for users, who may be unable to formulate search queries, have ill-defined search goals or may even struggle to understand search results. To ameliorate these difficulties, reinforcement learning-based information retrieval systems were developed to provide adaptive support to users. Reinforcement learning is used to build a model of user intent based on relevance feedback provided by the user. But how reliable is relevance feedback in this context? To answer this question, we developed a novel permutation-based metric for scoring the consistency of relevance feedback. We used this metric to perform a retrospective analysis of interaction data from lookup and exploratory search experiments. Our analysis shows that for lookup search relevance judgments are highly consistent, supporting previous findings that relevance feedback improves retrieval performance. For exploratory search, however, the distribution of consistency scores shows considerable inconsistency.

Any object-oriented action requires that the object be first brought into the attentional foreground, often through visual search. Outside the laboratory, this would always take place in the presence of a scene representation acquired from ongoing visual exploration. The interaction of scene memory with visual search is still not completely understood. Feature integration theory (FIT) has shaped both research on visual search, emphasizing the scaling of search times with set size when searches entail feature conjunctions, and research on visual working memory through the change detection paradigm. Despite its neural motivation, there is no consistently neural process account of FIT in both its dimensions. We propose such an account that integrates (1) visual exploration and the building of scene memory, (2) the attentional detection of visual transients and the extraction of search cues, and (3) visual search itself. The model uses dynamic field theory in which networks of neural dynamic populations supporting stable activation states are coupled to generate sequences of processing steps. The neural architecture accounts for basic findings in visual search and proposes a concrete mechanism for the integration of working memory into the search process. In a behavioral experiment, we address the long-standing question of whether both the overall speed and the efficiency of visual search can be improved by scene memory. We find both effects and provide model fits of the behavioral results. In a second experiment, we show that the increase in efficiency is fragile, and trace that fragility to the resetting of spatial working memory.

This paper proposed a Visual Exploratory Search Engine solution based on cloud computing environment. We focus on improving the traditional search engine by providing a novel framework for people to obtain reliable, personalization and graphical representation search results. The novel strategy is composed of three major parts: 1) Raw information Collect: using novel meta-search engine to quick obtain vast amounts of raw information, 2) Raw information Analysis and Indexing: automatic semantic similarity calculation and semantic link construction based on cloud computing environment, 3) User Interfaces: breakthrough the traditional result use list to represent, we use graphic to display the search result. In comparison with traditional lookup search methodologies, the proposed method is characterized with three main advantages. 1) It infers rich semantic relationships between the query and other related concepts from large-scale meta-search results from Google and Baidu search engines, and representing semantic relationships via graphs, 2) The exploratory search approach enables users to naturally and effectively explore adventure and discover knowledge in a rich information world, 3) And the novel search engine is based on a personalized model, and will provide different user with different search experience. We envision the above method, technology and tools have practical merit in some specific and appropriate application scenario.

Many people with vision loss have difficulty locating and identifying familiar objects. There is very little known about the retinal aspects of visual search with vision loss. While eye-trackers have provided information about eye-movements in visual search, only instruments that allow the assessment of visual stimuli position directly on the retina provide information about the retinal aspects of visual search. This in turn documents the relationship between visual search performance and retinal pathology in macular disease. Characteristics of the retinal locus for identification and characteristics of the movement of the retinal locus for identification (saccades) were studied to ascertain their relationship to decreased efficiency in visual search for observers with vision loss.

One hundred and twenty-seven undergraduate students variously performed a computerized version of the Gollin (1960) Incomplete Figures Test, the Mooney (1957) Test of Incomplete Face Perception, the Poppelreuter (1917) Overlapping Figures Test, and a visual search task. Performance of male subjects was superior to that of female subjects on the Mooney test but inferior on the visual search task. Correlation and regression analyses showed that the only significant predictor of Gollin test scores was latency to identify all items in the Overlapping Figures Test. There was no relationship between performances on the Gollin and Mooney tests or between Gollin or Mooney test performance and visual search latency. The Gollin and Mooney tests appear to access different perceptual processes, none of which is dependent on the efficiency of visual search.

In this article, we examine how emotional and perceptual stimulus factors influence visual search efficiency. In an initial task, we run a visual search task, using a large number of target/distractor emotion combinations. In two subsequent tasks, we then assess measures of perceptual (rated and computational distances) and emotional (rated valence, arousal and potency) stimulus properties. In a series of regression analyses, we then explore the degree to which target salience (the size of target/distractor dissimilarities) on these emotional and perceptual measures predict the outcome on search efficiency measures (response times and accuracy) from the visual search task. The results show that both emotional and perceptual stimulus salience contribute to visual search efficiency. The results show that among the emotional measures, salience on arousal measures was more influential than valence salience. The importance of the arousal factor may be a contributing factor to contradictory history of results within this field.

The efficiency of how people search for an item in visual search has, traditionally, been thought to depend on bottom-up or top-down guidance cues. However, recent research has shown that the rate at which people visually search through a display is also affected by cognitive strategies. In this study, we investigated the role of choice in visual search, by asking whether giving people a choice alters both preference for a cognitively neutral task and search behavior. Two visual search conditions were examined: one in which participants were given a choice of visual search task (the choice condition), and one in which participants did not have a choice (the no-choice condition). The results showed that the participants in the choice condition rated the task as both more enjoyable and likeable than did the participants in the no-choice condition. However, despite their preferences, actual search performance was slower and less efficient in the choice condition than in the no-choice condition (Exp. 1). Experiment 2 showed that the difference in search performance between the choice and no-choice conditions disappeared when central executive processes became occupied with a task-switching task. These data concur with a choice-impaired hypothesis of search, in which having a choice leads to more motivated, active search involving executive processes.

Some points of criticism against the idea that attentional selection is controlled by bottom-up processing were dispelled by the attentional window account. The attentional window account claims that saliency computations during visual search are only performed for stimuli inside the attentional window. Therefore, a small attentional window may avoid attentional capture by salient distractors because it is likely that the salient distractor is located outside the window. In contrast, a large attentional window increases the chances of attentional capture by a salient distractor. Large and small attentional windows have been associated with efficient (parallel) and inefficient (serial) search, respectively. We compared the effect of a salient color singleton on visual search for a shape singleton during efficient and inefficient search. To vary search efficiency, the nontarget shapes were either similar or dissimilar with respect to the shape singleton. We found that interference from the color singleton was larger with inefficient than efficient search, which contradicts the attentional window account. While inconsistent with the attentional window account, our results are predicted by computational models of visual search. Because of target-nontarget similarity, the target was less salient with inefficient than efficient search. Consequently, the relative saliency of the color distractor was higher with inefficient than with efficient search. Accordingly, stronger attentional capture resulted. Overall, the present results show that bottom-up control by stimulus saliency is stronger when search is difficult, which is inconsistent with the attentional window account.

A large number of previous studies have demonstrated that visual search on Web pages is impacted by many factors, such as colors, fonts, information forms and information layout. However, there are few studies about the effect of information quantity for visual search on Web pages. To investigate this issue, we designed two experiments using eye-tracking. The information quantity of Web pages in Experiment 2 was the double of that in Experiment 1. Results show that the search time in Experiment 2 is significantly longer than that in Experiment 1. From the heat maps of Experiment 2, most of the fixations were on the upper area of Web pages, especially on the upper half. Further analysis indicates that in Experiment 2, both of the fixation duration and fixation count on the upper half of Web pages are significantly larger than that on the lower half. These results suggest that the increase of information quantity on Web pages reduces the users' visual search efficiency and this is the impact of information overload on Web pages. The reason for this effect is that the range of the users' vision is limited and they pay more attention to the upper area of Web page, especially the upper half.

Search systems in online content platforms are typically biased toward a minority of highly consumed items, reflecting the most common user behavior of navigating toward content that is already familiar and popular. Query suggestions are a powerful tool to support query formulation and to encourage exploratory search and content discovery. However, classic approaches for query suggestions typically rely either on semantic similarity, which lacks diversity and does not reflect user searching behavior, or on a collaborative similarity measure mined from search logs, which suffers from data sparsity and is biased by highly popular queries. In this work, we argue that the task of query suggestion can be modelled as a link prediction task on a heterogeneous graph including queries and documents, enabling Graph Learning methods to effectively generate query suggestions encompassing both semantic and collaborative information. We perform an offline evaluation on an internal Spotify dataset of search logs and on two public datasets, showing that node2vec leads to an accurate and diversified set of results, especially on the large scale real-world data. We then describe the implementation in an instant search scenario and discuss a set of additional challenges tied to the specific production environment. Finally, we report the results of a large scale A/B test involving millions of users and prove that node2vec query suggestions lead to an increase in online metrics such as coverage (+1.42% shown search results pages with suggestions) and engagement (+1.21% clicks), with a specifically notable boost in the number of clicks on exploratory search queries (+9.37%).

This article presents a rational model developed under the distributed cognition framework that explains how social tags influence knowledge acquisition and adaptation in exploratory ill-defined information tasks. The model provides integrated predictions on the interactions among link selections, use and creation of tags, and the formation of mental categories. The model shows that the quality of tags not only influences search efficiency, but also the quality of mental categories formed during exploratory search. In addition, the model shows that aggregate regularities can be explained by microstructures of behavior that emerged from the adaptive assimilation of concepts and categories of multiple users through the social tagging system. The model has important implications on how collaborative systems could influence higher-level cognitive activities.

We have compared visual search under conditions that tend to isolate the magnocellular, parvocellular and koniocellular channels of the human visual system. We used isoluminant red-green stimuli that do not modulate short-wavelength sensitive (SWS) cones to isolate the parvocellular pathway, isoluminant SWS-cone isolating stimuli to stimulate only the koniocellular system and addition of small luminance contrasts to selectively activate the magnocellular pathway. We found that in the case of conjunction search, where attentional resources were required, the red-green (parvocellular) system can use accompanying small luminance (magnocellular) signals to improve visual search. On the other hand, when using SWS-cone isolating stimuli to selectively stimulate the blue-yellow (koniocellular) system, addition of similar luminance signals did not increase the efficiency of the serial visual search. The results indicate that S-cone signals may be processed in a separate pathway that does not get converging inputs from the magnocellular pathway. This is unlike the case with the red-green opponent system, which functions more synergistically with the magnocellular system.

This is an eye-tracking study examining the effects of image segmentation and target number on visual search performance. A two-way repeated-measures computer-based visual search test was used for data collection. Thirty students participated in the test, in which they were asked to search for all of the Landolt Cs in 80 arrays of closed rings. The dependent variables were search time, accuracy, fixation count, and average fixation duration. Our principal findings were that some of the segmentation methods significantly improved accuracy, and reduced search time, fixation count, and average fixation duration, compared with the no-segmentation condition. Increased target number was found to be associated with longer search time, lower accuracy, more fixations, and longer average fixation duration. Our study indicates that although visual search tasks with multiple targets are relatively difficult, the visual search accuracy and efficiency can potentially be improved with the aid of image segmentation.