Matching Items (1,035)
Filtering by
- All Subjects: Computer Science
- Resource Type: Text
- Status: Published
Description
The construction of many families of combinatorial objects remains a challenging problem. A t-restriction is an array where a predicate is satisfied for every t columns; an example is a perfect hash family (PHF). The composition of a PHF and any t-restriction satisfying predicate P yields another t-restriction also satisfying P with more columns than the original t-restriction had. This thesis concerns three approaches in determining the smallest size of PHFs.
Firstly, hash families in which the associated property is satisfied at least some number lambda times are considered, called higher-index, which guarantees redundancy when constructing t-restrictions. Some direct and optimal constructions of hash families of higher index are given. A new recursive construction is established that generalizes previous results and generates higher-index PHFs with more columns. Probabilistic methods are employed to obtain an upper bound on the optimal size of higher-index PHFs when the number of columns is large. A new deterministic algorithm is developed that generates such PHFs meeting this bound, and computational results are reported.
Secondly, a restriction on the structure of PHFs is introduced, called fractal, a method from Blackburn. His method is extended in several ways; from homogeneous hash families (every row has the same number of symbols) to heterogeneous ones; and to distributing hash families, a relaxation of the predicate for PHFs. Recursive constructions with fractal hash families as ingredients are given, and improve upon on the best-known sizes of many PHFs.
Thirdly, a method of Colbourn and Lanus is extended in which they horizontally copied a given hash family and greedily applied transformations to each copy. Transformations of existential t-restrictions are introduced, which allow for the method to be applicable to any t-restriction having structure like those of hash families. A genetic algorithm is employed for finding the "best" such transformations. Computational results of the GA are reported using PHFs, as the number of transformations permitted is large compared to the number of symbols. Finally, an analysis is given of what trade-offs exist between computation time and the number of t-sets left not satisfying the predicate.
Firstly, hash families in which the associated property is satisfied at least some number lambda times are considered, called higher-index, which guarantees redundancy when constructing t-restrictions. Some direct and optimal constructions of hash families of higher index are given. A new recursive construction is established that generalizes previous results and generates higher-index PHFs with more columns. Probabilistic methods are employed to obtain an upper bound on the optimal size of higher-index PHFs when the number of columns is large. A new deterministic algorithm is developed that generates such PHFs meeting this bound, and computational results are reported.
Secondly, a restriction on the structure of PHFs is introduced, called fractal, a method from Blackburn. His method is extended in several ways; from homogeneous hash families (every row has the same number of symbols) to heterogeneous ones; and to distributing hash families, a relaxation of the predicate for PHFs. Recursive constructions with fractal hash families as ingredients are given, and improve upon on the best-known sizes of many PHFs.
Thirdly, a method of Colbourn and Lanus is extended in which they horizontally copied a given hash family and greedily applied transformations to each copy. Transformations of existential t-restrictions are introduced, which allow for the method to be applicable to any t-restriction having structure like those of hash families. A genetic algorithm is employed for finding the "best" such transformations. Computational results of the GA are reported using PHFs, as the number of transformations permitted is large compared to the number of symbols. Finally, an analysis is given of what trade-offs exist between computation time and the number of t-sets left not satisfying the predicate.
ContributorsDougherty, Ryan Edward (Author) / Colbourn, Charles J (Thesis advisor) / Czygrinow, Andrzej (Committee member) / Forrest, Stephanie (Committee member) / Richa, Andrea (Committee member) / Arizona State University (Publisher)
Created2019
Description
In order to deploy autonomous multi-robot teams for humanitarian demining in Colombia, two key problems need to be addressed. First, a robotic controller with limited power that can completely cover a dynamic search area is needed. Second, the Colombian National Army (COLAR) needs to increase its science, technology and innovation (STI) capacity to help develop, build and maintain such robots. Using Thangavelautham's (2012, 2017) Artificial Neural Tissue (ANT) control algorithm, a robotic controller for an autonomous multi-robot team was developed. Trained by a simple genetic algorithm, ANT is an artificial neural network (ANN) controller with a sparse, coarse coding network architecture and adaptive activation functions. Starting from the exterior of open, basic geometric grid areas, computer simulations of an ANT multi-robot team with limited time steps, no central controller and limited a priori information, covered some areas completely in linear time, and other areas near completely in quasi-linear time, comparable to the theoretical cover time bounds of grid-based, ant pheromone, area coverage algorithms. To mitigate catastrophic forgetting, a new learning method for ANT, Lifelong Adaptive Neuronal Learning (LANL) was developed, where neural network weight parameters for a specific coverage task were frozen, and only the activation function and output behavior parameters were re-trained for a new coverage task. The performance of the LANL controllers were comparable to training all parameters ab initio, for a new ANT controller for the new coverage task.
To increase COLAR's STI capacity, a proposal for a new STI officer corps, Project ÉLITE (Equipo de Líderes en Investigación y Tecnología del Ejército) was developed, where officers enroll in a research intensive, master of science program in applied mathematics or physics in Colombia, and conduct research in the US during their final year. ÉLITE is inspired by the Israel Defense Forces Talpiot program.
To increase COLAR's STI capacity, a proposal for a new STI officer corps, Project ÉLITE (Equipo de Líderes en Investigación y Tecnología del Ejército) was developed, where officers enroll in a research intensive, master of science program in applied mathematics or physics in Colombia, and conduct research in the US during their final year. ÉLITE is inspired by the Israel Defense Forces Talpiot program.
ContributorsKwon, Byong (Author) / Castillo-Chavez, Carlos (Thesis advisor) / Thangavelautham, Jekanthan (Committee member) / Seshaiyer, Padmanabhan (Committee member) / Arizona State University (Publisher)
Created2019
Description
Parents fulfill a pivotal role in early childhood development of social and communication
skills. In children with autism, the development of these skills can be delayed. Applied
behavioral analysis (ABA) techniques have been created to aid in skill acquisition.
Among these, pivotal response treatment (PRT) has been empirically shown to foster
improvements. Research into PRT implementation has also shown that parents can be
trained to be effective interventionists for their children. The current difficulty in PRT
training is how to disseminate training to parents who need it, and how to support and
motivate practitioners after training.
Evaluation of the parents’ fidelity to implementation is often undertaken using video
probes that depict the dyadic interaction occurring between the parent and the child during
PRT sessions. These videos are time consuming for clinicians to process, and often result
in only minimal feedback for the parents. Current trends in technology could be utilized to
alleviate the manual cost of extracting data from the videos, affording greater
opportunities for providing clinician created feedback as well as automated assessments.
The naturalistic context of the video probes along with the dependence on ubiquitous
recording devices creates a difficult scenario for classification tasks. The domain of the
PRT video probes can be expected to have high levels of both aleatory and epistemic
uncertainty. Addressing these challenges requires examination of the multimodal data
along with implementation and evaluation of classification algorithms. This is explored
through the use of a new dataset of PRT videos.
The relationship between the parent and the clinician is important. The clinician can
provide support and help build self-efficacy in addition to providing knowledge and
modeling of treatment procedures. Facilitating this relationship along with automated
feedback not only provides the opportunity to present expert feedback to the parent, but
also allows the clinician to aid in personalizing the classification models. By utilizing a
human-in-the-loop framework, clinicians can aid in addressing the uncertainty in the
classification models by providing additional labeled samples. This will allow the system
to improve classification and provides a person-centered approach to extracting
multimodal data from PRT video probes.
skills. In children with autism, the development of these skills can be delayed. Applied
behavioral analysis (ABA) techniques have been created to aid in skill acquisition.
Among these, pivotal response treatment (PRT) has been empirically shown to foster
improvements. Research into PRT implementation has also shown that parents can be
trained to be effective interventionists for their children. The current difficulty in PRT
training is how to disseminate training to parents who need it, and how to support and
motivate practitioners after training.
Evaluation of the parents’ fidelity to implementation is often undertaken using video
probes that depict the dyadic interaction occurring between the parent and the child during
PRT sessions. These videos are time consuming for clinicians to process, and often result
in only minimal feedback for the parents. Current trends in technology could be utilized to
alleviate the manual cost of extracting data from the videos, affording greater
opportunities for providing clinician created feedback as well as automated assessments.
The naturalistic context of the video probes along with the dependence on ubiquitous
recording devices creates a difficult scenario for classification tasks. The domain of the
PRT video probes can be expected to have high levels of both aleatory and epistemic
uncertainty. Addressing these challenges requires examination of the multimodal data
along with implementation and evaluation of classification algorithms. This is explored
through the use of a new dataset of PRT videos.
The relationship between the parent and the clinician is important. The clinician can
provide support and help build self-efficacy in addition to providing knowledge and
modeling of treatment procedures. Facilitating this relationship along with automated
feedback not only provides the opportunity to present expert feedback to the parent, but
also allows the clinician to aid in personalizing the classification models. By utilizing a
human-in-the-loop framework, clinicians can aid in addressing the uncertainty in the
classification models by providing additional labeled samples. This will allow the system
to improve classification and provides a person-centered approach to extracting
multimodal data from PRT video probes.
ContributorsCopenhaver Heath, Corey D (Author) / Panchanathan, Sethuraman (Thesis advisor) / McDaniel, Troy (Committee member) / Venkateswara, Hemanth (Committee member) / Davulcu, Hasan (Committee member) / Gaffar, Ashraf (Committee member) / Arizona State University (Publisher)
Created2019
Description
Signal compressed using classical compression methods can be acquired using brute force (i.e. searching for non-zero entries in component-wise). However, sparse solutions require combinatorial searches of high computations. In this thesis, instead, two Bayesian approaches are considered to recover a sparse vector from underdetermined noisy measurements. The first is constructed using a Bernoulli-Gaussian (BG) prior distribution and is assumed to be the true generative model. The second is constructed using a Gamma-Normal (GN) prior distribution and is, therefore, a different (i.e. misspecified) model. To estimate the posterior distribution for the correctly specified scenario, an algorithm based on generalized approximated message passing (GAMP) is constructed, while an algorithm based on sparse Bayesian learning (SBL) is used for the misspecified scenario. Recovering sparse signal using Bayesian framework is one class of algorithms to solve the sparse problem. All classes of algorithms aim to get around the high computations associated with the combinatorial searches. Compressive sensing (CS) is a widely-used terminology attributed to optimize the sparse problem and its applications. Applications such as magnetic resonance imaging (MRI), image acquisition in radar imaging, and facial recognition. In CS literature, the target vector can be recovered either by optimizing an objective function using point estimation, or recovering a distribution of the sparse vector using Bayesian estimation. Although Bayesian framework provides an extra degree of freedom to assume a distribution that is directly applicable to the problem of interest, it is hard to find a theoretical guarantee of convergence. This limitation has shifted some of researches to use a non-Bayesian framework. This thesis tries to close this gab by proposing a Bayesian framework with a suggested theoretical bound for the assumed, not necessarily correct, distribution. In the simulation study, a general lower Bayesian Cram\'er-Rao bound (BCRB) bound is extracted along with misspecified Bayesian Cram\'er-Rao bound (MBCRB) for GN model. Both bounds are validated using mean square error (MSE) performances of the aforementioned algorithms. Also, a quantification of the performance in terms of gains versus losses is introduced as one main finding of this report.
ContributorsAlhowaish, Abdulhakim (Author) / Richmond, Christ D (Thesis advisor) / Papandreou-Suppappola, Antonia (Committee member) / Sankar, Lalitha (Committee member) / Arizona State University (Publisher)
Created2019
Description
The marked increase in the inflow of remotely sensed data from satellites have trans- formed the Earth and Space Sciences to a data rich domain creating a rich repository for domain experts to analyze. These observations shed light on a diverse array of disciplines ranging from monitoring Earth system components to planetary explo- ration by highlighting the expected trend and patterns in the data. However, the complexity of these patterns from local to global scales, coupled with the volume of this ever-growing repository necessitates advanced techniques to sequentially process the datasets to determine the underlying trends. Such techniques essentially model the observations to learn characteristic parameters of data-generating processes and highlight anomalous planetary surface observations to help domain scientists for making informed decisions. The primary challenge in defining such models arises due to the spatio-temporal variability of these processes.
This dissertation introduces models of multispectral satellite observations that sequentially learn the expected trend from the data by extracting salient features of planetary surface observations. The main objectives are to learn the temporal variability for modeling dynamic processes and to build representations of features of interest that is learned over the lifespan of an instrument. The estimated model parameters are then exploited in detecting anomalies due to changes in land surface reflectance as well as novelties in planetary surface landforms. A model switching approach is proposed that allows the selection of the best matched representation given the observations that is designed to account for rate of time-variability in land surface. The estimated parameters are exploited to design a change detector, analyze the separability of change events, and form an expert-guided representation of planetary landforms for prioritizing the retrieval of scientifically relevant observations with both onboard and post-downlink applications.
This dissertation introduces models of multispectral satellite observations that sequentially learn the expected trend from the data by extracting salient features of planetary surface observations. The main objectives are to learn the temporal variability for modeling dynamic processes and to build representations of features of interest that is learned over the lifespan of an instrument. The estimated model parameters are then exploited in detecting anomalies due to changes in land surface reflectance as well as novelties in planetary surface landforms. A model switching approach is proposed that allows the selection of the best matched representation given the observations that is designed to account for rate of time-variability in land surface. The estimated parameters are exploited to design a change detector, analyze the separability of change events, and form an expert-guided representation of planetary landforms for prioritizing the retrieval of scientifically relevant observations with both onboard and post-downlink applications.
ContributorsChakraborty, Srija (Author) / Papandreou-Suppappola, Antonia (Thesis advisor) / Christensen, Philip R. (Philip Russel) (Thesis advisor) / Richmond, Christ (Committee member) / Maurer, Alexander (Committee member) / Arizona State University (Publisher)
Created2019
Description
The traditional access control system suffers from the problem of separation of data ownership and management. It poses data security issues in application scenarios such as cloud computing and blockchain where the data owners either do not trust the data storage provider or even do not know who would have access to their data once they are appended to the chain. In these scenarios, the data owner actually loses control of the data once they are uploaded to the outside storage. Encryption-before-uploading is the way to solve this issue, however traditional encryption schemes such as AES, RSA, ECC, bring about great overheads in key management on the data owner end and could not provide fine-grained access control as well.
Attribute-Based Encryption (ABE) is a cryptographic way to implement attribute-based access control, which is a fine-grained access control model, thus solving all aforementioned issues. With ABE, the data owner would encrypt the data by a self-defined access control policy before uploading the data. The access control policy is an AND-OR boolean formula over attributes. Only users with attributes that satisfy the access control policy could decrypt the ciphertext. However the existing ABE schemes do not provide some important features in practical applications, e.g., user revocation and attribute expiration. Furthermore, most existing work focus on how to use ABE to protect cloud stored data, while not the blockchain applications.
The main objective of this thesis is to provide solutions to add two important features of the ABE schemes, i.e., user revocation and attribute expiration, and also provide a practical trust framework for using ABE to protect blockchain data. To add the feature of user revocation, I propose to add user's hierarchical identity into the private attribute key. In this way, only users whose identity is not revoked and attributes satisfy the access control policy could decrypt the ciphertext. To add the feature of attribute expiration, I propose to add the attribute valid time period into the private attribute key. The data would be encrypted by access control policy where all attributes have a temporal value. In this way, only users whose attributes both satisfy the access policy and at the same time these attributes do not expire,
are allowed to decrypt the ciphertext. To use ABE in the blockchain applications, I propose an ABE-enabled trust framework in a very popular blockchain platform, Hyperledger Fabric. Based on the design, I implement a light-weight attribute certificate authority for attribute distribution and validation; I implement the proposed ABE schemes and provide a toolkit which supports system setup, key generation,
data encryption and data decryption. All these modules were integrated into a demo system for protecting sensitive les in a blockchain application.
Attribute-Based Encryption (ABE) is a cryptographic way to implement attribute-based access control, which is a fine-grained access control model, thus solving all aforementioned issues. With ABE, the data owner would encrypt the data by a self-defined access control policy before uploading the data. The access control policy is an AND-OR boolean formula over attributes. Only users with attributes that satisfy the access control policy could decrypt the ciphertext. However the existing ABE schemes do not provide some important features in practical applications, e.g., user revocation and attribute expiration. Furthermore, most existing work focus on how to use ABE to protect cloud stored data, while not the blockchain applications.
The main objective of this thesis is to provide solutions to add two important features of the ABE schemes, i.e., user revocation and attribute expiration, and also provide a practical trust framework for using ABE to protect blockchain data. To add the feature of user revocation, I propose to add user's hierarchical identity into the private attribute key. In this way, only users whose identity is not revoked and attributes satisfy the access control policy could decrypt the ciphertext. To add the feature of attribute expiration, I propose to add the attribute valid time period into the private attribute key. The data would be encrypted by access control policy where all attributes have a temporal value. In this way, only users whose attributes both satisfy the access policy and at the same time these attributes do not expire,
are allowed to decrypt the ciphertext. To use ABE in the blockchain applications, I propose an ABE-enabled trust framework in a very popular blockchain platform, Hyperledger Fabric. Based on the design, I implement a light-weight attribute certificate authority for attribute distribution and validation; I implement the proposed ABE schemes and provide a toolkit which supports system setup, key generation,
data encryption and data decryption. All these modules were integrated into a demo system for protecting sensitive les in a blockchain application.
ContributorsDong, Qiuxiang (Author) / Huang, Dijiang (Thesis advisor) / Sen, Arunabha (Committee member) / Doupe, Adam (Committee member) / Arizona State University (Publisher)
Created2020
Description
The mobile crowdsensing (MCS) applications leverage the user data to derive useful information by data-driven evaluation of innovative user contexts and gathering of information at a high data rate. Such access to context-rich data can potentially enable computationally intensive crowd-sourcing applications such as tracking a missing person or capturing a highlight video of an event. Using snippets and pictures captured from multiple mobile phone cameras with specific contexts can improve the data acquired in such applications. These MCS applications require efficient processing and analysis to generate results in real time. A human user, mobile device and their interactions cause a change in context on the mobile device affecting the quality contextual data that is gathered. Usage of MCS data in real-time mobile applications is challenging due to the complex inter-relationship between: a) availability of context, context is available with the mobile phones and not with the cloud, b) cost of data transfer to remote cloud servers, both in terms of communication time and energy, and c) availability of local computational resources on the mobile phone, computation may lead to rapid battery drain or increased response time. The resource-constrained mobile devices need to offload some of their computation.
This thesis proposes ContextAiDe an end-end architecture for data-driven distributed applications aware of human mobile interactions using Edge computing. Edge processing supports real-time applications by reducing communication costs. The goal is to optimize the quality and the cost of acquiring the data using a) modeling and prediction of mobile user contexts, b) efficient strategies of scheduling application tasks on heterogeneous devices including multi-core devices such as GPU c) power-aware scheduling of virtual machine (VM) applications in cloud infrastructure e.g. elastic VMs. ContextAiDe middleware is integrated into the mobile application via Android API. The evaluation consists of overheads and costs analysis in the scenario of ``perpetrator tracking" application on the cloud, fog servers, and mobile devices. LifeMap data sets containing actual sensor data traces from mobile devices are used to simulate the application run for large scale evaluation.
This thesis proposes ContextAiDe an end-end architecture for data-driven distributed applications aware of human mobile interactions using Edge computing. Edge processing supports real-time applications by reducing communication costs. The goal is to optimize the quality and the cost of acquiring the data using a) modeling and prediction of mobile user contexts, b) efficient strategies of scheduling application tasks on heterogeneous devices including multi-core devices such as GPU c) power-aware scheduling of virtual machine (VM) applications in cloud infrastructure e.g. elastic VMs. ContextAiDe middleware is integrated into the mobile application via Android API. The evaluation consists of overheads and costs analysis in the scenario of ``perpetrator tracking" application on the cloud, fog servers, and mobile devices. LifeMap data sets containing actual sensor data traces from mobile devices are used to simulate the application run for large scale evaluation.
ContributorsPore, Madhurima (Author) / Gupta, Sandeep K. S. (Thesis advisor, Committee member) / Banerjee, Ayan (Committee member) / Reisslein, Martin (Committee member) / CERIN, CHRISTOPHE (Committee member) / Arizona State University (Publisher)
Created2019
Description
Robotic lower limb prostheses provide new opportunities to help transfemoral amputees regain mobility. However, their application is impeded by that the impedance control parameters need to be tuned and optimized manually by prosthetists for each individual user in different task environments. Reinforcement learning (RL) is capable of automatically learning from interacting with the environment. It becomes a natural candidate to replace human prosthetists to customize the control parameters. However, neither traditional RL approaches nor the popular deep RL approaches are readily suitable for learning with limited number of samples and samples with large variations. This dissertation aims to explore new RL based adaptive solutions that are data-efficient for controlling robotic prostheses.
This dissertation begins by proposing a new flexible policy iteration (FPI) framework. To improve sample efficiency, FPI can utilize either on-policy or off-policy learning strategy, can learn from either online or offline data, and can even adopt exiting knowledge of an external critic. Approximate convergence to Bellman optimal solutions are guaranteed under mild conditions. Simulation studies validated that FPI was data efficient compared to several established RL methods. Furthermore, a simplified version of FPI was implemented to learn from offline data, and then the learned policy was successfully tested for tuning the control parameters online on a human subject.
Next, the dissertation discusses RL control with information transfer (RL-IT), or knowledge-guided RL (KG-RL), which is motivated to benefit from transferring knowledge acquired from one subject to another. To explore its feasibility, knowledge was extracted from data measurements of able-bodied (AB) subjects, and transferred to guide Q-learning control for an amputee in OpenSim simulations. This result again demonstrated that data and time efficiency were improved using previous knowledge.
While the present study is new and promising, there are still many open questions to be addressed in future research. To account for human adaption, the learning control objective function may be designed to incorporate human-prosthesis performance feedback such as symmetry, user comfort level and satisfaction, and user energy consumption. To make the RL based control parameter tuning practical in real life, it should be further developed and tested in different use environments, such as from level ground walking to stair ascending or descending, and from walking to running.
This dissertation begins by proposing a new flexible policy iteration (FPI) framework. To improve sample efficiency, FPI can utilize either on-policy or off-policy learning strategy, can learn from either online or offline data, and can even adopt exiting knowledge of an external critic. Approximate convergence to Bellman optimal solutions are guaranteed under mild conditions. Simulation studies validated that FPI was data efficient compared to several established RL methods. Furthermore, a simplified version of FPI was implemented to learn from offline data, and then the learned policy was successfully tested for tuning the control parameters online on a human subject.
Next, the dissertation discusses RL control with information transfer (RL-IT), or knowledge-guided RL (KG-RL), which is motivated to benefit from transferring knowledge acquired from one subject to another. To explore its feasibility, knowledge was extracted from data measurements of able-bodied (AB) subjects, and transferred to guide Q-learning control for an amputee in OpenSim simulations. This result again demonstrated that data and time efficiency were improved using previous knowledge.
While the present study is new and promising, there are still many open questions to be addressed in future research. To account for human adaption, the learning control objective function may be designed to incorporate human-prosthesis performance feedback such as symmetry, user comfort level and satisfaction, and user energy consumption. To make the RL based control parameter tuning practical in real life, it should be further developed and tested in different use environments, such as from level ground walking to stair ascending or descending, and from walking to running.
ContributorsGao, Xiang (Author) / Si, Jennie (Thesis advisor) / Huang, He Helen (Committee member) / Santello, Marco (Committee member) / Papandreou-Suppappola, Antonia (Committee member) / Arizona State University (Publisher)
Created2020
Description
Machine learning (ML) and deep neural networks (DNNs) have achieved great success in a variety of application domains, however, despite significant effort to make these networks robust, they remain vulnerable to adversarial attacks in which input that is perceptually indistinguishable from natural data can be erroneously classified with high prediction confidence. Works on defending against adversarial examples can be broadly classified as correcting or detecting, which aim, respectively at negating the effects of the attack and correctly classifying the input, or detecting and rejecting the input as adversarial. In this work, a new approach for detecting adversarial examples is proposed. The approach takes advantage of the robustness of natural images to noise. As noise is added to a natural image, the prediction probability of its true class drops, but the drop is not sudden or precipitous. The same seems to not hold for adversarial examples. In other word, the stress response profile for natural images seems different from that of adversarial examples, which could be detected by their stress response profile. An evaluation of this approach for detecting adversarial examples is performed on the MNIST, CIFAR-10 and ImageNet datasets. Experimental data shows that this approach is effective at detecting some adversarial examples on small scaled simple content images and with little sacrifice on benign accuracy.
ContributorsSun, Lin (Author) / Bazzi, Rida (Thesis advisor) / Li, Baoxin (Committee member) / Tong, Hanghang (Committee member) / Arizona State University (Publisher)
Created2019
Description
Social networking sites like Twitter have provided people a platform to connect
with each other, to discuss and share information and news or to entertain themselves. As the number of users continues to grow there has been explosive growth in the data generated by these users. Such a vast data source has provided researchers a way to study and monitor public health.
Accurately analyzing tweets is a difficult task mainly because of their short length, the inventive spellings and creative language expressions. Instead of focusing at the topic level, identifying tweets that have personal health experience mentions would be more helpful to researchers, governments and other organizations. Another important limitation in the current systems for social media health applications is the use of a disease-specific model and dataset to study a particular disease. Identifying adverse drug reactions is an important part of the drug development process. Detecting and extracting adverse drug mentions in tweets can supplement the list of adverse drug reactions that result from the drug trials and can help in the improvement of the drugs.
This thesis aims to address these two challenges and proposes three systems. A generalizable system to identify personal health experience mentions across different disease domains, a system for automatic classifications of adverse effects mentions in tweets and a system to extract adverse drug mentions from tweets. The proposed systems use the transfer learning from language models to achieve notable scores on Social Media Mining for Health Applications(SMM4H) 2019 (Weissenbacher et al. 2019) shared tasks.
with each other, to discuss and share information and news or to entertain themselves. As the number of users continues to grow there has been explosive growth in the data generated by these users. Such a vast data source has provided researchers a way to study and monitor public health.
Accurately analyzing tweets is a difficult task mainly because of their short length, the inventive spellings and creative language expressions. Instead of focusing at the topic level, identifying tweets that have personal health experience mentions would be more helpful to researchers, governments and other organizations. Another important limitation in the current systems for social media health applications is the use of a disease-specific model and dataset to study a particular disease. Identifying adverse drug reactions is an important part of the drug development process. Detecting and extracting adverse drug mentions in tweets can supplement the list of adverse drug reactions that result from the drug trials and can help in the improvement of the drugs.
This thesis aims to address these two challenges and proposes three systems. A generalizable system to identify personal health experience mentions across different disease domains, a system for automatic classifications of adverse effects mentions in tweets and a system to extract adverse drug mentions from tweets. The proposed systems use the transfer learning from language models to achieve notable scores on Social Media Mining for Health Applications(SMM4H) 2019 (Weissenbacher et al. 2019) shared tasks.
ContributorsGondane, Shubham Bhagwan (Author) / Baral, Chitta (Thesis advisor) / Anwar, Saadat (Committee member) / Devarakonda, Murthy (Committee member) / Arizona State University (Publisher)
Created2019