無線通信和移動技術已經從根本上改變了人和人之間相互通信的方式，隨著像智能手機這樣功能強大的移動設備不斷普及，現在我們有更多的機會去監測用戶的運動狀態、社交情況和地理位置等信息。近期，越來越多的基於智能手機的傳感研究相繼出現，這些研究利用智能手機中的多種傳感、定位以及近距離無線設備來識別手機用戶當前的活動狀態和周圍環境。一些可識別用戶活動狀態和監控身體健康狀況的移動應用程式已經被開發并投入使用。儘管如此，當前大部份關於智能手機的研究忽視了這樣一個問題，智能手機是用戶與外界通信的一個指令中心。移動用戶可以使用智能手機用很多種方式聯繫他們的朋友，例如打電話、發送短消息、電子郵件、或者通過即時通信程序或者社交網絡，這些多渠道的通信方式和人與人之間面對面的交流一樣重要，因此智能手機是識別用戶和其他聯繫人的社會關係的關鍵。在本論文中，我們提出用智能手機中 獨有的多渠道用戶通信數據來對用戶的的社會關係進行分類。作為我們研究的開始，我們生成人工的通信數據並且用社交矩陣來為人與人之間的通信建立模型，這也幫助我們測試了很多可以應用在此類問題的數據挖掘算法。接下來，我們通過招募真實用戶來採集他們的各種社交通信數據，這些數據包括手機通話記錄、電子郵件、社交網絡（Facebook和Renren）和面對面的交流。通過在社交矩陣上應用不同的分類算法，我們發現SVM的分類性能要超過KNN和決策樹算法，SVM對於社會關係的分類準確率可以達到82.4%。我們也對來自不同渠道的通信數據進行了比較，最終發現來自社交網絡和面對面交流的數據在社交關係分類中起更大的作用。另外，我們通過使用降低維度算法可以把社交矩陣從65維度映射到9維度，關係分類的準確率卻沒有明顯降低，在降低維度的過程中我們也可以提取出用戶主要的通信特徵，從而更好地解釋社會關係分類的原理。最後，我們也應用了CUR矩陣分解算法從社交矩陣65列中選出13列建立新的社交矩陣，關係分類的準確率從82.4%降低到77.7%，但是我們卻可以通過 CUR來選擇合適的傳感器抽樣採集頻率，這樣可以在利用手機採集數據過程中節省更多手機電量。 / Wireless Communications and Mobile Computing have fundamentally changed the way people interact and communicate with each other. The proliferation of powerful and programmable mobile devices, smartphones in particular, has offered an unprecedented opportunity to continuously monitor the physical, social and geographical activities of their users. Recently, much research has been done on smartphone-based sensing which leverages the rich set of sensing, positioning and short-range radio capabilities of the smartphones to identify the context of user activities and ambient environment conditions. Mobile applications for personal behavior tracking and physical wellness monitoring have also been developed. Despite that, most of the existing work in mobile sensing has neglected the role of smartphone as the command-center of the user’s communications with the outside world. As mobile users contact their friends via phone, SMS, emails, instant messaging, and other online social-networking applications, these multi-modal communication activities are as equally important as physical activities in proling one’s life. They also hold the key to understand the user’s social relationship with other people of interest. In this thesis, we propose to use the unique multi-model interaction data from smartphone to classify social relationships. To jump start our study, we generate articial interaction data and build social interaction matrix to modeMl the interaction between people. This also helps us in testing a wide range of data mining analysis techniques for this type of problem. We then carry out a social interaction data collection campaign with a group of real users to obtain real-life multi-modal communication data, e.g., phone call, Email, online social network(Facebook and Renren), and physical location/proximity. After applying different classification algorithms on social interaction matrix, we find that SVM outperforms KNN and decision tree algorithms, with a classification accuracy of 82.4% (the accuracies of KNN and decision tree are 79.9% and 77.6%, respectively). We also compare the data from different interaction channels and finally find that on-line social network and location/proximity data contribute more to the overall classification accuracy. Additionally, with dimensionality reduction algorithms, the social interaction matrix can be embedded from 65 to 9 dimensional space while preserving the high classification accuracy and we also get principle interaction features as by-product. At last, we use CUR decomposi¬tion to select 13 out of the 65 features in the social interaction matrix. The classification accuracy drops from 82.4% to 77.7% after CUR decomposition. But it can help to determine the right sensor sampling frequency so as to enhance energy efficiency for social data collection. / Detailed summary in vernacular field only. / Sun, Deyi. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 90-96). / Abstracts also in Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Research Background --- p.7 / Chapter 2.1 --- Related work of social relationship analysis --- p.7 / Chapter 2.1.1 --- Community detection in social network --- p.8 / Chapter 2.1.2 --- Social influence analysis --- p.10 / Chapter 2.1.3 --- Modeling social interaction data --- p.10 / Chapter 2.1.4 --- Social relationship prediction --- p.12 / Chapter 2.2 --- Classification methodologies --- p.14 / Chapter 2.2.1 --- Algorithms for social relationship classification --- p.14 / Chapter 2.2.2 --- Algorithms for dimensionality reduction --- p.16 / Chapter 3 --- Problem Formulation of Relationship Classicification --- p.19 / Chapter 3.1 --- Multi-modal data in smartphones --- p.20 / Chapter 3.2 --- Formulation of relationship classification problem --- p.21 / Chapter 3.3 --- Refinement of feature definition and energy efficiency --- p.27 / Chapter 3.4 --- Chapter summary --- p.28 / Chapter 4 --- Social Interaction Data Acquisition --- p.30 / Chapter 4.1 --- Social interaction data collection campaign overview --- p.31 / Chapter 4.2 --- Format of raw interaction data --- p.33 / Chapter 4.3 --- Building social interaction matrix with real-life interaction data --- p.37 / Chapter 4.4 --- Chapter summary --- p.43 / Chapter 5 --- Statistical Analysis of Social Interaction Data --- p.45 / Chapter 5.1 --- Coverage of social interaction data --- p.46 / Chapter 5.2 --- Social relationships statistics --- p.48 / Chapter 5.3 --- Social relationship interaction patterns --- p.52 / Chapter 5.4 --- Chapter summary --- p.59 / Chapter 6 --- Automatic Social Relationship Classification Based on Smartphone Interaction Data --- p.61 / Chapter 6.1 --- Comparison of different classification algorithms --- p.62 / Chapter 6.2 --- Advantages of multi-modal interaction data --- p.65 / Chapter 6.3 --- Comparison of interaction data in different communication channels --- p.67 / Chapter 6.4 --- Dimensionality reduction on social interaction data --- p.72 / Chapter 6.5 --- Discussions in deploying social relationship classification application --- p.80 / Chapter 6.5.1 --- Considerations of user privacy --- p.81 / Chapter 6.5.2 --- Saving smartphone resources --- p.82 / Chapter 6.6 --- Chapter summary --- p.83 / Chapter 7 --- Conclusion and Future Work --- p.86 / Bibliography --- p.90

The Association for Computing Machinery (ACM) recognizes five computing disciplines: Computer Science (CS), Computer Engineering (CE), Information Technology (IT), Information Systems (IS), and Software Engineering (SE). Founded in 1947 the ACM is the world's largest society for computing educators, researchers, and professionals. While Computer Science has been a degree program since 1962, the other four are relatively new. This research focuses on understanding the graduate research in four of the five ACM disciplines (CS, CE, IT, and IS) using a large body of thesis and dissertation metadata. SE is not found in the metadata and graduate work in SE is not included. IS is no longer officially found in the metadata so its representative ProQuest replacement, Information Science although not an ACM recognized discipline is used based on the commonality of the associated ProQuest Classification code. The research is performed using co-word and graph analysis of author-supplied Classifications, Departments, and keywords. Similarities and differences between the disciplines are identified. Whether the computing discipline is the primary or the secondary focus of the research makes a large difference in the connections it makes with other academic disciplines. It was found that the Departments from which computing research originates varies widely but the majority come from computing-related Departments. Finally, gaps are apparent from the practitioners' views of the computing disciplines versus the public's view.

With the current trend in cloud computing and virtualization, more organizations are moving their systems from a physical host to a virtual server. Although this can significantly reduce hardware, power, and administration costs, it can increase the cost of analyzing performance problems. With virtualization, there is an initial performance overhead, and as more virtual machines are added to a physical host the interference increases between various guest machines. When this interference occurs, a virtualized guest application may not perform as expected. There is little or no information to the virtual OS about the interference, and the current performance tools in the guest are unable to show this interference. We examine the interference that has been shown in previous research, and relate that to existing tools and research in root cause analysis. We show that in virtualization there are additional layers which need to be analyzed, and design a framework to determine if degradation is occurring from an external virtualization layer. Additionally, we build a virtualization test suite with Xen and PostgreSQL and run multiple tests to create I/O interference. We show that our method can distinguish between a problem caused by interference from external systems and a problem from within the virtual guest.

Indiana University-Purdue University Indianapolis (IUPUI) / The contribution of this thesis focuses on addressing the challenges of improving and integrating the UniFrame Discovery Service (URDS) and Multi-level Matching (MLM) concepts. The objective was to find enhancements for both URDS and MLM and address the need of a comprehensive discovery service which goes beyond simple attribute based matching. It presents a detailed discussion on developing an enhanced version of URDS with MLM (proURDS). After implementing proURDS, the thesis includes details of experiments with different deployments of URDS components and different configurations of MLM. The experiments and analysis were carried out using proURDS produced MLM contracts. The proURDS referred to a public dataset called QWS dataset. This dataset includes actual information of software components (i.e., web services), which were harvested from the Internet. The proURDS implements the different matching operations as independent operators at each level of matching (i.e., General, Syntactic, Semantic, Synchronization, and QoS). Finally, a case study was carried out with the deployed proURDS. The case study addresses real world component discovery requirements from the earth science domain. It uses the contracts collected from public portals which provide geographical and weather related data.

The use of optical input/output (I/O) interconnects, in addition to electrical I/Os, is a promising approach for achieving high-bandwidth, chip-to-board communications required for future high-performance gigascale chip-based systems. While numerous efforts are underway to investigate the integration of optoelectronics and silicon microelectronics, virtually no work has been reported relating to testing of such chips. The objective of this research is to explore methods that enable wafer-level testing of gigascale chips having electrical and optical I/O interconnects. A major challenge in achieving this is to develop probe modules which would allow high-precision, temporary interconnection of a multitude of electrical and optical I/Os, in a chip-size area, to automated test equipment. A probe module would need to do this in a rapid, step-and-repeat manner across all the chips on the wafer. In this work, two candidate probe modules were devised, batch-fabricated on Si using microfabrication techniques, and successfully demonstrated. The first probe module consists of compliant electrical probes (10^3 probes/cm^2) fabricated alongside grating-in-waveguide optical probes. The second module consists of micro-opto-electro-mechanical-systems (MOEMS)-based microsocket probes (10^4 probes/cm^2) to interface a chip with polymer pillar-based electrical and optical I/Os. High-density through-wafer interconnects are an essential attribute in both probe substrates for transferring electrical and optical signals to the substrate back-side. Fabrication and characterization of metal-clad, metal-filled, and polymer-filled through-wafer interconnects as well as process integration with probe substrate fabrication are described and numerous possible redistribution schemes are explicated. Chips with optical and electrical I/Os are an emerging technology, and one that test engineers are likely to encounter in the near future. The contributions of this thesis are to help understand and address the issues relating to joint electrical and optical testing during manufacturing.

Indiana University-Purdue University Indianapolis (IUPUI) / The amount of information produced in the form of electronic free text in healthcare is increasing to levels incapable of being processed by humans for advancement of his/her professional practice. Information extraction (IE) is a sub-field of natural language processing with the goal of data reduction of unstructured free text. Pertinent to IE is an annotated corpus that frames how IE methods should create a logical expression necessary for processing meaning of text. Most annotation approaches seek to maximize meaning and knowledge by chunking sentences into phrases and mapping these phrases to a knowledge source to create a logical expression. However, these studies consistently have problems addressing semantics and none have addressed the issue of semantic similarity (or synonymy) to achieve data reduction. To achieve data reduction, a successful methodology for data reduction is dependent on a framework that can represent currently popular phrasal methods of IE but also fully represent the sentence. This study explores and reports on the benefits, problems, and requirements to using the predicate-argument statement (PAS) as the framework. A convenient sample from a prior study with ten synsets of 100 unique sentences from radiology reports deemed by domain experts to mean the same thing will be the text from which PAS structures are formed.

Indiana University-Purdue University Indianapolis (IUPUI) / Proteins are nano-machines that work inside every living organism. Functional disruption of one or several proteins is the cause for many diseases. However, the functions for most proteins are yet to be annotated because inexpensive sequencing techniques dramatically speed up discovery of new protein sequences (265 million and counting) and experimental examinations of every protein in all its possible functional categories are simply impractical. Thus, it is necessary to develop computational function-prediction tools that complement and guide experimental studies. In this study, we developed a series of predictors for highly accurate prediction of proteins with DNA-binding, RNA-binding and carbohydrate-binding capability. These predictors are a template-based technique that combines sequence and structural information with predicted binding affinity. Both sequence and structure-based approaches were developed. Results indicate the importance of binding affinity prediction for improving sensitivity and precision of function prediction. Application of these methods to the human genome and structure genome targets demonstrated its usefulness in annotating proteins of unknown functions and discovering moon-lighting proteins with DNA,RNA, or carbohydrate binding function. In addition, we also investigated disruption of protein functions by naturally occurring genetic variations due to insertions and deletions (INDELS). We found that protein structures are the most critical features in recognising disease-causing non-frame shifting INDELs. The predictors for function predictions are available at http://sparks-lab.org/spot, and the predictor for classification of non-frame shifting INDELs is available at http://sparks-lab.org/ddig.

Indiana University-Purdue University Indianapolis (IUPUI) / Mobile IP though allows mobility features to a node it suffers from signaling Latencies which are mainly incurred due to the fact that the MN itself is involved in the handover process. To overcome this problem proxy mobile IPv6(PMIPv6) was defined where the mobility signaling is taken care of by a proxy server while keeping track of the MN's movement. PMIPv6 has considerably reduced the handover latency but the demand for real time applications over the network has increased tremendously due to recent explosion of the cloud era. My thesis focuses on increasing the L3 handoff signaling efficiency by reducing the latency. This is achieved by our idea to do both the AAA authentication as well as the LMA registration in PMIPv6 at the same time. The simulation results show that our proposed approach perform better than the current PMIPv6 L3 handover signaling reducing the latency as well as packet loss.

Indiana University-Purdue University Indianapolis (IUPUI) / There have been many endeavors to design humanoid robots that have human characteristics such as dexterity, autonomy and intelligence. Humanoid robots are intended to cooperate with humans and perform useful work that humans can perform. The main advantage of humanoid robots over other machines is that they are flexible and multi-purpose. In this thesis, a human-like robotic arm is designed and used in a task which is typically performed by humans, namely, catching a ball. The robotic arm was designed to closely resemble a human arm, based on anthropometric studies. A rigid multibody dynamics software was used to create a virtual model of the robotic arm, perform experiments, and collect data. The inverse kinematics of the robotic arm was solved using a Newton-Raphson numerical method with a numerically calculated Jacobian. The system was validated by testing its ability to find a kinematic solution for the catch position and successfully catch the ball within the robot's workspace. The tests were conducted by throwing the ball such that its path intersects different target points within the robot's workspace. The method used for determining the catch location consists of finding the intersection of the ball's trajectory with a virtual catch plane. The hand orientation was set so that the normal vector to the palm of the hand is parallel to the trajectory of the ball at the intersection point and a vector perpendicular to this normal vector remains in a constant orientation during the catch. It was found that this catch orientation approach was reliable within a 0.35 x 0.4 meter window in the robot's workspace. For all tests within this window, the robotic arm successfully caught and dropped the ball in a bin. Also, for the tests within this window, the maximum position and orientation (Euler angle) tracking errors were 13.6 mm and 4.3 degrees, respectively. The average position and orientation tracking errors were 3.5 mm and 0.3 degrees, respectively. The work presented in this study can be applied to humanoid robots in industrial assembly lines and hazardous environment recovery tasks, amongst other applications.

Indiana University-Purdue University Indianapolis (IUPUI) / Adaptive optics-optical coherence tomography (AO-OCT) is a non-invasive method of imaging the human retina in vivo. It can be used to visualize microscopic structures, making it incredibly useful for the early detection and diagnosis of retinal disease. The research group at Indiana University has a novel multi-camera AO-OCT system capable of 1 MHz acquisition rates. Until this point, a method has not existed to process data from such a novel system quickly and accurately enough on a CPU, a GPU, or one that can scale to multiple GPUs automatically in an efficient manner. This is a barrier to using a MHz AO-OCT system in a clinical environment. A novel approach to processing AO-OCT data from the unique multi-camera optics system is tested on multiple graphics processing units (GPUs) in parallel with one, two, and four camera combinations. The design and results demonstrate a scalable, reusable, extensible method of computing AO-OCT output. This approach can either achieve real time results with an AO-OCT system capable of 1 MHz acquisition rates or be scaled to a higher accuracy mode with a fast Fourier transform of 16,384 complex values.