arXiv subjects

Enhong Chen

Publications and source records attributed to Enhong Chen.

335 records · Page 19Linked to original sources

arXiv2015

Information Coverage Maximization in Social Networks

Social networks, due to their popularity, have been studied extensively these years. A rich body of these studies is related to influence maximization, which aims to select a set of seed nodes for maximizing the expected number of active nodes at the end of the process. However, the set of active nodes can not fully represent the true coverage of information propagation. A node may be informed of the information when any of its neighbours become active and try to activate it, though this node (namely informed node) is still inactive. Therefore, we need to consider both active nodes and informed nodes that are aware of the information when we study the coverage of information propagation in a network. Along this line, in this paper we propose a new problem called Information Coverage Maximization that aims to maximize the expected number of both active nodes and informed ones. After we prove that this problem is NP-hard and submodular in the independent cascade model and the linear threshold model, we design two algorithms to solve it. Extensive experiments on three real-world data sets demonstrate the performance of the proposed algorithms.

cs.SI↗

arXiv2015

Learning Better Word Embedding by Asymmetric Low-Rank Projection of Knowledge Graph

Word embedding, which refers to low-dimensional dense vector representations of natural words, has demonstrated its power in many natural language processing tasks. However, it may suffer from the inaccurate and incomplete information contained in the free text corpus as training data. To tackle this challenge, there have been quite a few works that leverage knowledge graphs as an additional information source to improve the quality of word embedding. Although these works have achieved certain success, they have neglected some important facts about knowledge graphs: (i) many relationships in knowledge graphs are \emph{many-to-one}, \emph{one-to-many} or even \emph{many-to-many}, rather than simply \emph{one-to-one}; (ii) most head entities and tail entities in knowledge graphs come from very different semantic spaces. To address these issues, in this paper, we propose a new algorithm named ProjectNet. ProjecNet models the relationships between head and tail entities after transforming them with different low-rank projection matrices. The low-rank projection can allow non \emph{one-to-one} relationships between entities, while different projection matrices for head and tail entities allow them to originate in different semantic spaces. The experimental results demonstrate that ProjectNet yields more accurate word embedding than previous works, thus leads to clear improvements in various natural language processing tasks.

cs.CL↗

arXiv2014

Agent Behavior Prediction and Its Generalization Analysis

Machine learning algorithms have been applied to predict agent behaviors in real-world dynamic systems, such as advertiser behaviors in sponsored search and worker behaviors in crowdsourcing. The behavior data in these systems are generated by live agents: once the systems change due to the adoption of the prediction models learnt from the behavior data, agents will observe and respond to these changes by changing their own behaviors accordingly. As a result, the behavior data will evolve and will not be identically and independently distributed, posing great challenges to the theoretical analysis on the machine learning algorithms for behavior prediction. To tackle this challenge, in this paper, we propose to use Markov Chain in Random Environments (MCRE) to describe the behavior data, and perform generalization analysis of the machine learning algorithms on its basis. Since the one-step transition probability matrix of MCRE depends on both previous states and the random environment, conventional techniques for generalization analysis cannot be directly applied. To address this issue, we propose a novel technique that transforms the original MCRE into a higher-dimensional time-homogeneous Markov chain. The new Markov chain involves more variables but is more regular, and thus easier to deal with. We prove the convergence of the new Markov chain when time approaches infinity. Then we prove a generalization bound for the machine learning algorithms on the behavior data generated by the new Markov chain, which depends on both the Markovian parameters and the covering number of the function class compounded by the loss function for behavior prediction and the behavior prediction model. To the best of our knowledge, this is the first work that performs the generalization analysis on data generated by complex processes in real-world dynamic systems.

cs.LG↗

arXiv2013

Query Segmentation for Relevance Ranking in Web Search

In this paper, we try to answer the question of how to improve the state-of-the-art methods for relevance ranking in web search by query segmentation. Here, by query segmentation it is meant to segment the input query into segments, typically natural language phrases, so that the performance of relevance ranking in search is increased. We propose employing the re-ranking approach in query segmentation, which first employs a generative model to create top $k$ candidates and then employs a discriminative model to re-rank the candidates to obtain the final segmentation result. The method has been widely utilized for structure prediction in natural language processing, but has not been applied to query segmentation, as far as we know. Furthermore, we propose a new method for using the result of query segmentation in relevance ranking, which takes both the original query words and the segmented query phrases as units of query representation. We investigate whether our method can improve three relevance models, namely BM25, key n-gram model, and dependency model. Our experimental results on three large scale web search datasets show that our method can indeed significantly improve relevance ranking in all the three cases.

cs.IR↗

arXiv2013

Cell Oscillation Resolution in Mobility Profile Building

This paper has been withdrawn by the authors. Mobility profile building became extensively examined area in Location based services (LBS) through extraction of significant locations. Mobility traces are recorded under three reference positioning systems that are Satellite based i.e. GPS, Network based i.e. GSM and Local positioning i.e. WLAN, RFID, IrDA. Satellite based and local positioning due to of high power consumption, additional resource installation, low accuracy and space limitation are less encouraging. So network based positioning i.e. GSM is only viable solution for mobility tracing through Cell global identity (CGI). CGI presents the Cell-ids to extract the significant locations from mobility history. However CGI faces cell oscillation problem, where user is assigned multiple Cell-Ids even at a stationary state for load balancing and GSM cells overlapping. In this paper we proposed two semi-supervised methodology for cell oscillation resolution i.e. semantic tagging and overlapped area clustering, the proposed methodologies are equally useful for the identification of significant places too.

cs.NI↗

arXiv2012

Spatial Outlier Detection from GSM Mobility Data

This paper has been withdrawn by the authors. With the rigorous growth of cellular network many mobility datasets are available publically, which attracted researchers to study human mobility fall under spatio-temporal phenomenon. Mobility profile building is main task in spatio-temporal trend analysis which can be extracted from the location information available in the dataset. The location information is usually gathered through the GPS, service provider assisted faux GPS and Cell Global Identity (CGI). Because of high power consumption and extra resource installation requirement in GPS related methods, Cell Global Identity is most inexpensive method and readily available solution for location information. CGI location information is four set head i.e. Mobile country code (MCC), Mobile network code (MNC), Location area code (LAC) and Cell ID, location information is retrieved in form of longitude and latitude coordinates through any of publically available Cell Id databases e.g. Google location API using CGI. However due to of fast growth in GSM network, change in topology by the GSM service provider and technology shift toward 3G exact spatial extraction is somehow a problem in it, so location extraction must dealt with spatial outlier's problem first for mobility building. In this paper we proposed a methodology for the detection of spatial outliers from GSM CGI data, the proposed methodology is hierarchical clustering based and used the basic GSM network architecture properties.

cs.NI↗

arXiv2012

Precise Location Acquisition of Mobility Data Using Cell-id

This paper has been withdrawn by the authors. Cellular network data has become a hot source of study for extraction of user-mobility and spatio-temporal trends. Location binding in mobility data can be done through different methods like GPS, service provider assisted faux-GPS and Cell Global Identity (CGI). Among these Cell Global Identity is most inexpensive method and readily available solution for mobility extraction; however exact spatial extraction is somehow a problem in it. This paper presents the spatial extraction technique of mobile phone user raw data which carries the information like location information, proximity location and activity of subjects. This work mainly focuses on the data pre-processing methodology and technique to interpret the low level mobility data into high level mobility information using the designed clustering methodology and publically available Cell-IDs databases. Work proposed the semi- supervised strategy to derive the missing locations thorough the usage of semantic tag information and removal of spatial outliers for precise mobility profile building.

cs.NI↗

arXiv2012

Enterprise Resource Planning - Real blessing or a Blessing in Disguise : An Exploration of the Contextual Factors in Public Sector

Information systems have always been in a prime focus in organizations in both local (Pakistani) and global environment. Now the race of being the best through Information Systems has created its importance in public sector organizations to meet the global challenges. Public sector organizations have been facing problems in different segments of technology adoption especially in ERP projects. ERP adoption/implementation projects in public sector organizations still encounter major setbacks in terms of partly/completely success/failure. Cultural and other social barriers have been resistant in technology adoption in Pakistan. Now in the case of big ERP adoptions the contextual factors must be identified and addressed. The paper investigates the reasons of success or failure by addressing nature of complexities regarding different contextual factors. The study includes a sample of Pakistan s four public sector organizations. The sample of this four organizations includes two organizations (Type-A) i.e. Oil & Gas Development Company Limited (OGDCL) and National Database Registration Authority (NADRA) where ERP has been successfully implemented and other two (Type-B) i.e. Pakistan Telecommunication Corporation Limited (PTCL), Higher Education Commission (HEC) where ERP implementation is in progress. The findings address the contextual factors i.e. cultural, environmental & political changes which have a variable impact on ERP systems adoption/implementation in addition to Business Process Re-engineering (BPR). Paper also briefly includes analysis of gaps between pre & post ERP implementation scenarios.

cs.CY↗

arXiv2012

Performance Enhancement Factors of ERP Projects in a Telecom Public Sector Organization of Pakistan : An Exploratory Study

Public sector organizations are treated in a different manner, as Information technology/information system has become necessity in a highly competitive environment. Importance of information systems is becoming more and more vital as the global technology adoption is in progress. ERP projects are considered as one of the most important and critical area of technology especially in public sector organizations where cost effectiveness and operational efficiency is prioritized on profits. A lot of research studies have been made on ERP projects but mostly on critical success factors (CSFs) and other managerial issues. This ongoing research mainly focuses upon the performance of ERP software in public sector organizations by thoroughly going through one of the largest most public sector organizations of Pakistan. Though ERP projects are handled by experienced consultants in addition to the support of vendor companies but because of its features and functional complexity and mismatch with the organizational processes, different managerial and technical issues arise during and after its implementation. This study investigates the performance of ERP system in a large public sector organization of Pakistan; five most critical technical factors, which can lead the whole project towards success or failure, have been dug out. This exploratory study i.e. extensive literature review and a case study includes survey and interviews which later on investigated through their application in a public sector organization in Pakistan.

cs.CY↗

arXiv2012

A Linear Circuit Model For Social Influence Analysis

Understanding the behaviors of information propagation is essential for the effective exploitation of social influence in social networks. However, few existing influence models are both tractable and efficient for describing the information propagation process and quantitatively measuring social influence. To this end, in this paper, we develop a linear social influence model, named Circuit due to its close relation to the circuit network. Based on the predefined four axioms of social influence, we first demonstrate that our model can efficiently measure the influence strength between any pair of nodes. Along this line, an upper bound of the node(s)' influence is identified for potential use, e.g., reducing the search space. Furthermore, we provide the physical implication of the Circuit model and also a deep analysis of its relationships with the existing methods, such as PageRank. Then, we propose that the Circuit model provides a natural solution to the problems of computing each single node's authority and finding a set of nodes for social influence maximization. At last, the effectiveness of the proposed model is evaluated on the real-world data. The extensive experimental results demonstrate that Circuit model consistently outperforms the state-of-the-art methods and can greatly alleviate the computation burden of the influence maximization problem.

cs.SI↗

arXiv2012

A Social Influence Model Based On Circuit Theory

Understanding the behaviors of information propagation is essential for the effective exploitation of social influence in social networks. However, few existing influence models are tractable and efficient for describing the information propagation process, especially when dealing with the difficulty of incorporating the effects of combined influences from multiple nodes. To this end, in this paper, we provide a social influence model that alleviates this obstacle based on electrical circuit theory. This model vastly improves the efficiency of measuring the influence strength between any pair of nodes, and can be used to interpret the real-world influence propagation process in a coherent way. In addition, this circuit theory model provides a natural solution to the social influence maximization problem. When applied to realworld data, the circuit theory model consistently outperforms the state-of-the-art methods and can greatly alleviate the computation burden of the influence maximization problem.

cs.SI↗