274 research outputs found
Multi-theme sentiment analysis with sentiment shifting
No full textBusiness reviews contain rich sentiment on multiple themes, disclosing more interesting information than the overall polarities of documents. When it comes to fine-grained sentiment analysis, given any segment of text, we are not only interested in overall polarity of such segment, but also the sentiment words play major effects. However, sentiment analysis at the word level poses significant challenges due to the complexity of reviews, the inconsistency of sentiment in different themes, and the sentiment shifting resulting from linguistic patterns---contextual valence shifters.
To simultaneously resolve the multi-theme and sentiment shifting dilemma, a unified explainable sentiment analysis model, MTSA, is proposed in this paper, which enables both classification of sentiment polarity and discovery of quantified sentiment-shifting patterns. MTSA formulates multi-theme sentiment by learning embeddings (i.e., vector representations) for both themes and words, and derives the shifter effect learning algorithm by modeling the shifted sentiment in a logistic regression model.
Extensive experiments have been conducted on Yelp business reviews and IMDB movie reviews. The improvement of sentiment polarity classification demonstrates the effectiveness of MTSA at rectifying word feature representations of reviews, and the human evaluation shows its successful discovery of multi-theme sentiment words and automatic effect quantification of contextual valence shifters.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2018-05-01The student, Hongkun Yu, accepted the attached license on 2016-04-20 at 13:24.The student, Hongkun Yu, submitted this Thesis for approval on 2016-04-20 at 13:27.This Thesis was approved for publication on 2016-04-21 at 08:52.DSpace SAF Submission Ingestion Package generated from Vireo submission #9349 on 2016-07-07 at 13:50:14Made available in DSpace on 2016-07-07T20:27:48Z (GMT). No. of bitstreams: 2
YU-THESIS-2016.pdf: 1377626 bytes, checksum: fe33b7692b15580351e718e199b44e59 (MD5)
LICENSE.txt: 4207 bytes, checksum: 6ecb6be099535d4749440bbf2fcd7721 (MD5)
Previous issue date: 2016-04-21Embargo set by: Seth Robbins for item 93151
Lift date: 2018-07-07T20:28:14Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 93151
Lift date: 2018-07-07T20:35:34Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction Lifted for Item 93151 on 2018-07-08T09:15:16Z
Recommended from our members
Olfactory Evidence Accumulation in Mice
No full textIn nature, odor cues from distant objects are sparse and highly fluctuating due to turbulent airflow. Animals may integrate odor concentration sampled over time rather than rely on transient odor concentration to effectively locate an object. To study how animals integrate and weigh discrete olfactory evidence over time, I developed a new behavioral task in which mice make binary decisions under fluctuating odor stimuli over many seconds. A custom-built device allowed the precise delivery of discrete, short pulses of odors at arbitrary Poisson-distributed pulse rates. I found that trained mice can readily differentiate stochastic odor stimuli with different average pulse rates presented over many seconds. In order to investigate how active, discrete sniff-based sampling of a stochastically varying environmental cue affects the neural representation and perceptual interpretation of the cue, calcium imaging in the axon terminals of olfactory sensory neurons (OSNs) in the glomeruli of olfactory bulb (OB) was performed. I discovered that OSN activity was highly modulated by the phase of the sniffing cycle. Regression of behavioral outcome against the timing of odor pulses in the breathing cycle revealed a kernel that weighted pulses arriving during the inhalation cycle more than during exhalation. This kernel matched the OSN activity kernel over breathing cycle, suggesting that the strength of the perception elicited by single pulses was directly related to the strength of the OSN responses. Decision noise scaled with the number of pulses presented. Tetrode recordings of single-unit neural activities in the anterior piriform cortex (APC) showed high correlations with transient odor pulses, but not the accumulated evidence. The neural activities in APC exhibited diverse dependency on the phase of sniffing, ranging from being strongly modulated by the sniff cycles to sniff-cycle invariant. My study indicates that mice integrate discrete olfactory inputs over several seconds to make decisions and that perceptual evidence is weighted by the intensity of the OSN response to the input. Furthermore, the platform described in this dissertation introduces a new paradigm in perceptual decision-making in which I can, unlike in vision or audition, record neural activity at all levels, from the first layer of sensory neurons to the decision-making networks
Recommended from our members
Complementary Metal-Oxide-Semiconductor (CMOS) Bio-electronic Interface for Cell-based Phenotypic Drug Screening
No full textMicroelectrode Array (MEA) has been widely researched and with some commercial usage in measuring electrical properties and behavior of cardiac and neuronal networks, thanks to its low cost and parallelism compared to traditional patch clamp. Integration of Complementary Metal-Oxide-Semiconductor (CMOS) technology with MEA allows further miniaturization of the electrodes, improvement on the spatial resolution and the signal-to-noise ratio and enabling highly-paralleled real-time measurement and stimulation. CMOS-MEA, therefore, becomes an excellent research tool for in vitro electrophysiology studies, usually with a single-well device to measure and perform manual electrical stimulation on one cell culture at a time. Furthermore, CMOS-MEA also improves the multi-parametric label-free readouts on general cells from commercially available MEAs, opening the potential applications into phenotypic drug discovery.
This dissertation discusses the design and development of two CMOS-MEA Integrated Circuit (IC) systems. First, an extension of two previously published CMOS nanoelectrode array (CNEA) systems, the 3rd generation of CNEA, is presented. The 3rd generation of CNEA features 1,024 pixels capable of simultaneously recording, on-chip action potential detection, on-chip inter-pixel feedback decision, and arbitrary stimulation pattern generation. The on-chip action potential detection and inter-pixel feedback decision making allow a stimulation pattern to be generated and applied to an arbitrary pixel within a microsecond from an action potential detected on another pixel. This capability makes the 3rd generation CNEA a perfect candidate for Spike-Timing-Dependent-Plasticity (STDP) studies in neuronal networks.
The majority focus of this dissertation would be on the multiwell version of the CNEA (Multiwell Platform). The Multiwell Platform contains 24 custom design ICs interlinked onto a custom interposer printed circuit board (PCB), creating 96 identical wells and their associated circuitries into a standard form factor wellplate. Each well contains 4,096 pixels and 256 readout channels capable of scanning through the entire pixel array with arbitrary square patterns. The unique design and functionalities of the Multiwell Platform enable parallel measurements on multiple biological-relevant parameters on general cells and therefore enable high-throughput phenotypic drug screening applications
Morphology Modelling for On-line Wear Debris Monitoring
No full textOn-line Wear Debris Analysis (WDA) with its rapid fault detection capability in a non-destructive manner, has been increasingly employed in machine condition monitoring. However, the available debris features obtained from current on-line WDA processes are mostly restricted to statistical indicators due to the limited data quality. Morphological features extraction from individual debris, hence, has not been fully successful in practice. Moreover, wear debris detections are currently
relying on single-side or two-dimensional (2-D) debris features, while the three-dimensional (3-D) features that contain valuable morphological information are still not accessible in on-line WDA and its further application is thus impeded.
To extract more individual debris features in current on-line WDA, this research aims at formulating an efficient scheme with improved debris observation, robust features extraction and 3-D debris shape measurement. First, an automatic framework is established to collect debris information in multi-views by observing the moving debris in a flow cell. Since the debris profiles are captured when it is moving, motion blur inevitably occurs and degrades the image quality. A fusion based restoration method is then developed to remove the motion blur by integrating multiple deblurring processes over several localised Point Spread Function (PSF). Furthermore, out-of-focus blur often occurs as another predominant source of degradation that leads to low feature extraction accuracy. This problem is tackled here by utilising a Convolutional Neural Networks (CNN) to model the defocus process and then remove the degradation. Finally, given the improved debris profiles, a debris shape measurement procedure is constructed to estimate 3-D debris features by minimising the discrepancies between multiple potential reconstructions.
Through validation experiments and comparisons, it indicates that the proposed 3-D measurement framework could enable the observed debris information to be extended into the third dimension that is rarely achieved by available on-line detectors for WDA. Compared with other debris imaging approaches, the developed method allows the estimation of material loss based on the measured debris volume. Furthermore, the classification of debris shape into chunk, laminar and sphere is now accomplishable with no specialised equipment. The industrial practicability of WDA can be improved considerably
The trend of carbon emission in Henan Province under the background of renewable energy development
No full textA novel top-down fabrication process for Ge2Sb2Te5 phase change material nanowires
No full textA novel e-beam free, top-down spacer etch process was used to fabricate sub-hundred nanometer Ge2Sb2Te5 phase change nanowires. Naowires with a cross-section dimension of 50 nm × 100 nm (width × height) were obtained and phase change functionality demonstrated
Medium-term electric power demand forecasting based on economic-electricity transmission model
No full textResearch and implementation of power acquisition fault recognition system based on data mining
No full text- …
