Telecommunication has a common problem of suffering from high churn rate, e.g. an existing customer may leave the service provider or shift to another. Churn prediction is the task of predicting the likelihood that a certain group of customers churn, and customer segmentation has been widely applied for a target marketing strategy. Previous work has adopted various clustering and classification methods, among which support vector machines and neural networks have achieved better performance, but the possible reason why customers churn is rarely explored in details. This article aims to address the gap of understanding customer churn using customers’ usage of services. To do this, an integrated framework containing two phases was built: we carried out predictive analysis using churn prediction and association analysis (phase one), based on which we provided recommendations in the decision analysis (phase two). The experimental results showed that the proposed Usage-based Decision Tree (UsageDT) achieved better performance in understanding the churners/non-churners in terms of interpretability, higher precision, and much higher true positive rate. The usage-based association analysis was integrated seamlessly with churn prediction, and deepened our understanding of the churn phenomena. Evaluation carried out with two professionals in telecommunication industry also revealed the usefulness of our recommendations.

With the acceleration of urbanization and modern civilization, more and more complex regions are formed in urban area. Although understanding these regions could provide huge insights to facilitate valuable applications for urban planning and business intelligence, few methods have been developed to effectively capture the rapid transformation of urban regions. In recent years, the widely applied location-acquisition technologies offer a more effective way to capture the dynamics of a city through analyzing people's movement activities based on mobility data. However, several challenges exist, including data sparsity and difficulties in result understanding and validation. To tackle these challenges, in this paper, we propose MobiSeg, an interactive visual analytics system, which supports the exploration of people's movement activities to segment the urban area into regions sharing similar activity patterns. A joint analysis is conducted on three types of heterogeneous mobility data (i.e., taxi trajectories, metro passenger RFID card data, and telco data), which can complement each other and provide a full picture of people's activities in a region. In addition, advanced analytical algorithms (e.g., non-negative matrix factorization (NMF) based method to capture latent activity patterns, as well as metric learning to calibrate and supervise the underlying analysis) and novel visualization designs are integrated into our system to provide a comprehensive solution to region segmentation in urban areas. We demonstrate the effectiveness of our system via case studies with real-world datasets and qualitative interviews with domain experts.

The advanced telecom technologies and massive volumes of intelligent mobile phone users have yielded a huge amount of real-time data of people’s all-in-one telecommunication records, which we call telco big data. With telco data and the domain knowledge of an urban city, we are now able to analyze the movement and contact patterns of humans in an unprecedented scale. Flow map is widely used to display the movements of humans from one single source to multiple destinations by representing locations as nodes and movements as edges. However, it fails the task of visualizing both movement and contact data. In addition, analysts often need to compare and examine the patterns side by side, and do various quantitative analysis. In this work, we propose a novel spatio-temporal flow map layout to visualize when and where people from different locations move into the same places and make contact. We also propose integrating the spatiotemporal flow maps into existing spatiotemporal visualization techniques to form a suite of techniques for visualizing the movement and contact patterns. We report a potential application the proposed techniques can be applied to. The results show that our design and techniques properly unveil hidden information, while analysis can be achieved efficiently.

Understanding co-occurrence in urban human mobility (i.e. people from two regions visit an urban place during the same time span) is of great value in a variety of applications, such as urban planning, business intelligence, social behavior analysis, as well as containing contagious diseases. In recent years, the widespread use of mobile phones brings an unprecedented opportunity to capture large-scale and fine-grained data to study co-occurrence in human mobility. However, due to the lack of systematic and efficient methods, it is challenging for analysts to carry out in-depth analyses and extract valuable information. In this paper, we present TelCoVis, an interactive visual analytics system, which helps analysts leverage their domain knowledge to gain insight into the co-occurrence in urban human mobility based on telco data. Our system integrates visualization techniques with new designs and combines them in a novel way to enhance analysts' perception for a comprehensive exploration. In addition, we propose to study the correlations in co-occurrence (i.e. people from multiple regions visit different places during the same time span) by means of biclustering techniques that allow analysts to better explore coordinated relationships among different regions and identify interesting patterns. The case studies based on a real-world dataset and interviews with domain experts have demonstrated the effectiveness of our system in gaining insights into co-occurrence and facilitating various analytical tasks.

We show that telco big data can make churn prediction much more easier from the $3$V's perspectives: Volume, Variety, Velocity. Experimental results confirm that the prediction performance has been significantly improved by using a large volume of training data, a large variety of features from both business support systems (BSS) and operations support systems (OSS), and a high velocity of processing new coming data. We have deployed this churn prediction system in one of the biggest mobile operators in China. From millions of active customers, this system can provide a list of prepaid customers who are most likely to churn in the next month, having $0.96$ precision for the top $50000$ predicted churners in the list. Automatic matching retention campaigns with the targeted potential churners significantly boost their recharge rates, leading to a big business value.

With the increasing prevalence of versatile mobile devices and the fast deployment of broadband mobile networks, a huge volume of Mobile Broadband (MBB) data has been generated over time. The MBB data naturally contain rich information of a large number of mobile users, covering a considerable fraction of whole population nowadays, including the mobile applications they are using at different locations and time; the MBB data may present the unprecedentedly large knowledge base of human behavior which has highly recognized commercial and social value. However, the storage, management and analysis of the huge and fast growing volume of MBB data post new and significant challenges to the industrial practitioners and research community. In this demonstration, we present a new, MBB data tailored, distributed analytic system named OceanST which has addressed a series of problems and weaknesses of the existing systems, originally designed for more general purpose and capable to handle MBB data to some extent. OceanST is featured by ( i ) efficiently loading of ever-growing MBB data, ( ii ) a bunch of spatiotemporal aggregate queries and basic analysis APIs frequently found in various MBB data application scenarios, and ( iii ) sampling-based approximate solution with provable accuracy bound to cope with huge volume of MBB data. The demonstration will show the advantage of OceanST in a cluster of 5 machines using 3TB data.

UNLABELLED: Sequencing reads generated by RNA-sequencing (RNA-seq) must first be mapped back to the genome through alignment before they can be further analyzed. Current fast and memory-saving short-read mappers could give us a quick view of the transcriptome. However, they are neither designed for reads that span across splice junctions nor for repetitive reads, which can be mapped to multiple locations in the genome (multi-reads). Here, we describe a new software package: ABMapper, which is specifically designed for exploring all putative locations of reads that are mapped to splice junctions or repetitive in nature. AVAILABILITY AND IMPLEMENTATION: The software is freely available at: http://abmapper.sourceforge.net/. The software is written in C++ and PERL. It runs on all major platforms and operating systems including Windows, Mac OS X and LINUX.

BACKGROUND: RNA sequencing (RNA-seq) measures gene expression levels and permits splicing analysis. Many existing aligners are capable of mapping millions of sequencing reads onto a reference genome. For reads that can be mapped to multiple positions along the reference genome (multireads), these aligners may either randomly assign them to a location, or discard them altogether. Either way could bias downstream analyses. Meanwhile, challenges remain in the alignment of reads spanning across splice junctions. Existing splicing-aware aligners that rely on the read-count method in identifying junction sites are inevitably affected by sequencing depths. RESULTS: The distance between aligned positions of paired-end (PE) reads or two parts of a spliced read is dependent on the experiment protocol and gene structures. We here proposed a new method that employs an empirical geometric-tail (GT) distribution of intron lengths to make a rational choice in multireads selection and splice-sites detection, according to the aligned distances from PE and sliced reads. CONCLUSIONS: GT models that combine sequence similarity from alignment, and together with the probability of length distribution, could accurately determine the location of both multireads and spliced reads.

RNA sequencing (RNA-Seq) measures gene expression levels and permits splicing analysis. There are many aligners that provide ultra-fast mapping of millions of sequencing reads onto a reference genome. However, random assignment or removal of reads matching multiple positions along the reference genome could bias downstream analyses. Meanwhile, challenges remain in the alignment of reads spanning across splice junctions. Existing splicing-aware aligners that rely on the read-count method in identifying junction sites are inevitably affected by sequencing depths. We here proposed a new method that employs an empirical geometric-tail (GT) distribution of intron lengths to make a rational choice in multireads selection and junction sites detection.

In this work, we define generalized (sequence) patterns, which is based on several real Biological problems, including transcription factors (TFs) binding to transcription factor binding sites (TFBSs), cis-regulatory modules, protein domain analysis, and alternative splicing etc. Simply speaking, a generalized pattern is composed of several substrings with gaps in-between two substrings. We propose a generalized pattern matching algorithm that uses a complementary dualseeding strategy, which is sensitive to errors (both mismatches and indels). We also develop a generalized pattern matching tool, which is to our knowledge the first ever developed specially for generalized pattern matching. Rather than replacing the existing general purpose matching tools, such as BLAST, BLAT, and PatternHunter etc, our tool provides an alternative and helps users to solve real problems, especially those that can be modeled as generalized patterns. We use data randomly sampled from reference sequences of human genome (NCBI build v18) in experiments, and hit 98.74% generalized patterns on average. The tool runs on both LINUX and Windows platforms, and the memory peak goes to a little bit larger than 1GB only.

This work proposes a novel algorithm to do approximate matching in a database consisting of multiple sequences. We apply Agrep algorithm in an indexing structure, the r-cut numerical substring array (r-NSA). The structure basically indexes all the substrings of length r. The advantage of using the r-NSA is two-fold: (1) The space requirement of the r-NSA is much smaller than that of the other existing indexing structures, such as the generalized suffix tree. (2) We propose an algorithm to apply Agrep in the r-NSA, in which the substrings are processed sequentially. Since the common substrings are processed only once, the cost of our algorithm is smaller than that of the full scanning search by Agrep. Consequently, the matching time of our algorithm is also reduced. We design experiments to validate and compare the performance of our algorithm against the full scanning search by Agrep. We define the speed-up of our algorithm as the time required by the full scanning search by Agrep over that of our algorithm. We use eight sets of real DNA sequences in our experiments, and the results show that our algorithm achieves significant speed-up. We also investigate the speed-up of difference data sets, and analyze their differences in detail.