ACM TIST Special Issue on Visual Analytics
Most machine learning methods focus on classifying instances whose classes have already been seen in training. In practice, many applications require classifying instances whose classes have not been seen previously. Zero-shot learning is a powerful and promising learning paradigm, in which the classes covered by training instances and the classes we aim to classify are disjoint. In this paper, we provide a comprehensive survey of zero-shot learning. First of all, we provide a formal definition of zero-shot learning and introduce related concepts. According to the data utilized in model optimization, we classify zero-shot learning into three learning settings. Secondly, we describe different semantic spaces adopted in existing zero-shot learning works. Thirdly, we categorize existing zero-shot learning methods, introduce methods in each category, and discuss their advantages and disadvantages. Fourthly, we discuss different applications of zero-shot learning. Finally, we highlight future research directions on zero-shot learning from both the problem setting and the technique perspectives.
Understanding real-world event participation behavior has been a subject of active research and can offer valuable insights for event-related recommendation and advertisement. The emergence of event-based social networks (EBSNs), which attract online users to host/attend offline events, has enabled exciting new research in this domain. However, most existing works focus on understanding or predicting individual users' event participation behavior or recommending events to individual users. Few study has addressed the problem of event popularity from the event organizer's point of view. In this work, we study the latent factors for determining event popularity using large-scale datasets collected from the popular Meetup.com EBSN in five major cities around the world. We analyze and model four contextual factors: spatial factor using location convenience, quality, popularity density and competitiveness; group factor using group member entropy and loyalty; temporal factor using temporal preference and weekly event patterns; and semantic factor using readability, sentiment, part-of-speech and text novelty. In addition, we have developed a group-based social influence propagation network to model group-specific influences on events. By combining the COntextual features and Social Influence NEtwork, our integrated prediction framework COSINE can capture the diverse influential factors of event participation and can be used by event organizers to predict/improve the popularity of their events. Detailed evaluations demonstrate that our COSINE framework achieves high accuracy for event popularity prediction in all five cities with diverse cultures and user event behaviors.
We address the problem of exploring and comparing large collections of scored, directed networks for understanding inferred Bayesian networks used in biology. In this field, heuristic algorithms explore the space of possible network solutions, sampling this space based on algorithm parameters and a network score that encodes the statistical fit to the data. The goal of the analyst is to guide the heuristic search and decide how to determine a final consensus network structure, usually by selecting the top scoring network or constructing the consensus network from a collection of high scoring networks. BayesPiles, our visualisation tool, helps with understanding the structure of the solution space and supporting the construction of a final consensus network that is representative of the underlying data set. BayesPiles builds upon and extends MultiPiles to meet our domain requirements. We developed BayesPiles in conjunction with computational biologists who have used this tool on data sets used in their research. The biologists found our solution provides them with new insights and helps them achieve results that are representative of the underlying data.
The increased availability of large-scale trajectory data provides rich information for the study of urban dynamics. For example, New York City Taxi & Limousine Commission regularly releases source/destination information of taxi trips, where 173 million taxi trips released for Year 2013. Such a big dataset provides us potential new perspectives to address the traditional traffic problems. In this paper, we study the travel time estimation problem. Instead of following the traditional route-based travel time estimation, we propose to simply use a large amount of taxi trips without using the intermediate trajectory points to estimate the travel time between source and destination. Our experiments show very promising results. The proposed big data-driven approach significantly outperforms both state-of-the-art route-based method and online map services. Our study indicates that novel simple approaches could be empowered by the big data and these approaches could serve as new baselines for some traditional computational problems.
This paper aims to develop a new and robust approach to feature representation. Motivated by the success of Auto-Encoders, we frst theoretically analyze and summarize the general properties of all algorithms that are based on traditional Auto-Encoders: 1) The reconstruction error of the input can not be lower than a lower bound, which can be viewed as a guiding principle for reconstructing the input. Additionally, when the input is corrupted with noises, the reconstruction error of the corrupted input also can not be lower than a lower bound. 2) The reconstruction of a hidden representation achieving its ideal situation is the necessary condition for the reconstruction of the input to reach the ideal state. 3) Minimizing the Frobenius norm of the Jacobian matrix of the hidden representation has a defciency and may result in a much worse local optimum value. We believe that minimizing the reconstruction error of the hidden representation is more robust than minimizing the Frobenius norm of the Jacobian matrix of the hidden representation. Based on the above analysis, we propose a new model termed Double Denoising Auto-Encoders (DDAEs), which uses corruption and reconstruction on both the input and the hidden representation. We demonstrate that the proposed model is highly exible and extensible and has a potentially better capability to learn invariant and robust feature representations. We also show that for dealing with noises or inessential features, our model is more robust than Denoising Auto-Encoders (DAEs). Furthermore, we will detail how to train DDAEs with two di erent pre-training methods by optimizing the objective function in a combined and separate manner, respectively. Comparative experiments illustrate that the proposed model is signifcantly better for representation learning than the state-of-the-art models.
The paper provides new techniques for optimizing domain design for goal and plan recognition using plan libraries. We define two new problems: Goal Recognition Design for Plan Libraries (GRD-PL) and Plan Recognition Design (PRD). Solving the GRD-PL helps to infer which goal the agent is trying to achieve, while solving PRD can help to infer how the agent is going to achieve its goal. For each problem, we define a worst-case distinctiveness measure that is an upper bound on the number of observations that are necessary to unambiguously recognize the agent's goal or plan. The paper studies the relationship between these measures, showing that the worst-case distinctiveness of GRD-PL is a lower bound of the worst-case plan distinctiveness of PRD, and that they are equal under certain conditions. We provide two complete algorithms for minimizing the worst-case distinctiveness of plan libraries without reducing the agent's ability to complete its goals: One is a brute force search over all possible plans and one a constraint-based search that identifies plans that are most difficult to distinguish in the domain. These algorithms are evaluated in three hierarchical plan recognition settings from the literature. We were able to reduce the worst case distinctiveness of the domains using our approach, in some cases reaching 100% improvement within a predesignated time window. Our iterative algorithm outperforms the brute force approach by an order of magnitude in terms of runtime.
At the heart of multi-agent systems is the ability to cooperate in order to improve the performance of individual agents and/or the system as a whole. While a widespread assumption in the literature is that such cooperation is essentially unrestricted, in many realistic settings this assumption does not hold. A highly-influential approach for modelling such scenarios are graph-restricted games introduced by Myerson. In this approach, agents are represented by nodes in a graph, edges represent communication channels, and a group can generate an arbitrary value only if there exists a direct or indirect communication channel between every pair of agents within the group. Two fundamental solution concepts that were proposed for such games are the Myerson value and the Shapley value. While an algorithm has been developed to compute the Shapley value in arbitrary graph-restricted games, no such general-purpose algorithm has been developed for the Myerson value to date. With this in mind, we set to develop for such games a general-purpose algorithm to compute the Myerson value, and a more efficient algorithm to compute the Shapley value. Since the computation of either value involves enumerating all connected induced subgraphs of the games underlying graph, we start by developing an algorithm dedicated to this enumeration, and show empirically that it is faster than the state of the art in the literature. Finally, we present a sample application of both algorithms, in which we test the Myerson value and the Shapley value as advanced measures of node centrality in networks.
Massive public resume data emerging on the Internet indicates individual-related characteristics in terms of profile and career experiences. Resume Analysis (RA) provides opportunities for many applications, such as recruitment trend predict, talent seeking and evaluation. Existing RA studies either largely rely on the knowledge of domain experts, or leverage classic statistical or data mining models to identify and filter explicit attributes based on pre-defined rules. However, they fail to discover the latent semantic information from semi-structured resume text, i.e., individual career progress trajectory and social-relations, which are otherwise vital to comprehensive understanding of peoples career evolving patterns. Besides, when dealing with massive resumes, how to properly visualize such semantic information to reduce the information load and to support better human cognition is also challenging. To tackle these issues, we propose a visual analytics system ResumeVis to mine and visualize resume data. Firstly, a text-mining based approach is presented to extract semantic information. Then, a set of visualizations are devised to represent the semantic information in multiple perspectives. By interactive exploration on ResumeVis performed by domain experts, the following tasks can be accomplished: to trace individual career evolving trajectory; to mine latent social-relations among individuals; and to hold the full picture of massive resumes' collective mobility. Case studies with over 2500 online officer resumes demonstrate the effectiveness of our system.
As mobile devices are becoming ubiquitous nowadays, the geographical activities and interactions of human beings can be easily recorded and accessed. Each mobile individual can belong to an online social network. Unfortunately, the underlying online social relationships are hidden and only available to service providers. Acquiring the social network of mobile users would enrich lots of mobile applications, such as friend recommendation and energy-saving mobile database management. In this work, we propose to infer online social ties using purely offline geographical activities of users, such as check-in records and spatial meeting events. To tackle the problem, we devise a novel inference framework, O2O-Inf, which consists of two components, Feature Modeling and Link Inference. Feature modeling is to characterize both direct and indirect geographical interactions between nodes from co-location and graph features. Link inference aims to infer the social ties based on a small set of observed social links, and the idea is that pairs of nodes sharing similar geographical behaviors have the same tendency of linkage (i.e., either being friends or non-friends). Experiments conducted on a Gowalla location-based social network and a Meetup event-based social network exhibit satisfying performance in comparison to state-of-the-art prediction methods under the settings of offline-to-online network inference and geo-link prediction.