Sparsity Constraint Nonnegative Tensor Factorization for Mobility Pattern Mining

2020 IEEE Symposium Series on Computational Intelligence (SSCI)

Bashar

2020

Self Cite

Traditional topic modelling methods such as Latent Dirichlet Allocation (LDA) [2] are challenged with short text data and face sparsity and low word co-occurrence issues. Factorization methods like Non-negative Matrix Factorization (NMF) [3] map the high-dimensional (sparse) text representation to a lower-dimensional representation. These methods have become popular in text mining due to their capability to capture the patterns in the lower dimensional representation of the data [4]. NMF decomposes a high-dimensional (tweet × term) matrix into two low-rank factor matrices that represent tweet and term clusters. It produces a part-based representation by allowing only additive combinations of basis components [4].Non-negative Tensor Factorization (NTF) [5], an extension of NMF for high-dimensional data, can identify associations among multiple dimensions [6]. This brings an added advantage to NTF over NMF as the patterns can be interpreted with associations. NTF based methods have been used for spatiotemporal patterns elicitation on traditional (i.e. structured) data. NTF for spatio-temporal pattern elicitation on social media (i.e. unstructured) data brings multiple challenges. Firstly, representing this unstructured text data and spatiotemporal information in a single tensor representation model is challenging when there is a need to preserve the association among them. Secondly, the short texts from social media like Twitter can induce sparseness to the tensor representation. The state-of-the-art factorization algorithms may fail to effectively learn the spatio-temporal patterns in the presence of noise and sparsity present in social media data [7], [8]. Finally, the larger data size introduces efficiency issues in factorization process [9]. Therefore, a proper data representation and selection of a suitable factorization algorithm is crucial to deal with social media data.In this paper, we present a novel NTF based spatio-temporal topic dynamics discovery method. We focus on applying the best-suited data representation model and the factorization algorithm to understand the spatio-temporal distribution of topics emerging from users' interactions on social media related to Covid-19. We present a case study analysing a large tweet Abstract-Social media platforms facilitate mankind a datadriven world by enabling billions of people to share their thoughts and activities ubiquitously. This huge collection of data, if analysed properly, can provide useful insights into people's behavior. More than ever, now is a crucial time under the Covid-19 pandemic to understand people's online behaviors detailing what topics are being discussed, and where (space) and when (time) they are discussed. Given the high complexity and poor quality of the huge social media data, an effective spatio-temporal topic detection method is needed. This paper proposes a tensor-based representation of social media data and Non-negative Tensor Factorization (NTF) to identify the topics discussed in social media data along with the spatio-tempora...

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Existing Data Models: Matrix Representationmentioning

confidence: 99%

See 2 more Smart Citations

Understanding the Spatio-temporal Topic Dynamics of Covid-19 using Nonnegative Tensor Factorization: A Case Study

2020 IEEE Symposium Series on Computational Intelligence (SSCI)

Bashar

2020

Self Cite

“…With the advent of tagging, sensor and Internet of Things (IoT) technologies, online interaction data can be easily generated with tagging or spatio-temporal information. Tensor models become the natural choice to represent a multi-dimensional dataset, for example as (user × imaдe × taд) or (user × location × time) [7,18,25,44]. A Tensor Factorization (TF) based method decomposes the tensor model into multiple factor matrices where each matrix learns the latent features inherent in the usage dataset [25,56].…”

Section: Introductionmentioning

confidence: 99%

Efficient Nonnegative Tensor Factorization via Saturating Coordinate Descent

ACM Trans. Knowl. Discov. Data

Yuen

2020

Self Cite

With the advancements in computing technology and web-based applications, data are increasingly generated in multi-dimensional form. These data are usually sparse due to the presence of a large number of users and fewer user interactions. To deal with this, the Nonnegative Tensor Factorization (NTF) based methods have been widely used. However existing factorization algorithms are not suitable to process in all three conditions of size, density, and rank of the tensor. Consequently, their applicability becomes limited. In this article, we propose a novel fast and efficient NTF algorithm using the element selection approach. We calculate the element importance using Lipschitz continuity and propose a saturation point-based element selection method that chooses a set of elements column-wise for updating to solve the optimization problem. Empirical analysis reveals that the proposed algorithm is scalable in terms of tensor size, density, and rank in comparison to the relevant state-of-the-art algorithms.

Identifying Covid-19 misinformation tweets and learning their spatio-temporal topic dynamics using Nonnegative Coupled Matrix Tensor Factorization

Luong

et al. 2021

Soc. Netw. Anal. Min.

Self Cite

Social media platforms like Twitter have become an easy portal for billions of people to connect and exchange their thoughts. Unfortunately, people commonly use these platforms to share misinformation which can influence other people adversely. The spread of misinformation is unavoidable in an extraordinary situation like Covid-19, and the consequences can be dreadful. This paper proposes a two-step ranking-based misinformation detection (RMiD) technique. Firstly, a novel ranking-based approach leveraging the scalable information retrieval infrastructure is applied to detect misinformation from a huge collection of unlabelled tweets based on a related but very small labelled misinformation data set. Secondly, the identified misinformation tweets are represented as a coupled matrix tensor model and Nonnegative Coupled Matrix Tensor Factorization is applied to learn their spatio-temporal topic dynamics. The experimental analysis shows that RMiD is capable of detecting misinformation with better coverage and less noise in comparison with existing techniques. Moreover, the coupled matrix tensor representation has improved the quality of topics discovered from unlabelled data up to 4% by leveraging the semantic similarity of terms in labelled data. Supplementary Information The online version supplementary material available at 10.1007/s13278-021-00767-7.