Automatic Labeling of Topic Models Using Graph-Based Ranking

He, Dongbin; Wang, Minjuan; Khattak, Abdul Mateen; Zhang, Li; Gao, Wanlin

doi:10.1109/access.2019.2940516

Cited by 12 publications

(4 citation statements)

References 37 publications

(77 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results of the proposed method showed that the use of summaries as labels had apparent advantages compared with the use of words and phrases. He et al [45] built a graph-based ranking model, namely TLRank, that aimed to automatically label every topic produced from topic models. The proposed model used the strategy of enhancing matrix transition probability based on the textual similarity between vertices and the characteristics of vertices (sentences).…”

Section: Automatic Naming Of Topicsmentioning

confidence: 99%

Musawah: A Data-Driven AI Approach and Tool to Co-Create Healthcare Services with a Case Study on Cancer Disease in Saudi Arabia

et al. 2022

View full text Add to dashboard Cite

The sustainability of human existence is in dire danger and this threat applies to our environment, societies, and economies. Smartization of cities and societies has the potential to unite individuals and nations towards sustainability as it requires engaging with our environments, analyzing them, and making sustainable decisions regulated by triple bottom line (TBL). Poor healthcare systems affect individuals, societies, the planet, and economies. This paper proposes a data-driven artificial intelligence (AI) based approach called Musawah to automatically discover healthcare services that can be developed or co-created by various stakeholders using social media analysis. The case study focuses on cancer disease in Saudi Arabia using Twitter data in the Arabic language. Specifically, we discover 17 services using machine learning from Twitter data using the Latent Dirichlet Allocation algorithm (LDA) and group them into five macro-services, namely, Prevention, Treatment, Psychological Support, Socioeconomic Sustainability, and Information Availability. Subsequently, we show the possibility of finding additional services by employing a topical search over the dataset and have discovered 42 additional services. We developed a software tool from scratch for this work that implements a complete machine learning pipeline using a dataset containing over 1.35 million tweets we curated during September–November 2021. Open service and value healthcare systems based on freely available information can revolutionize healthcare in manners similar to the open-source revolution by using information made available by the public, the government, third and fourth sectors, or others, allowing new forms of preventions, cures, treatments, and support structures.

show abstract

Section: Automatic Naming Of Topicsmentioning

confidence: 99%

Musawah: A Data-Driven AI Approach and Tool to Co-Create Healthcare Services with a Case Study on Cancer Disease in Saudi Arabia

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Multiple applications utilise the relative information to get the desired outcome. Authors in [25] created a graph-based ranking method for labeling the topic from the terms belonging to the topic and Kullback-Leibler Divergence to select the relevant candidate sentences. The authors created a directed weighted graph using relevance centrality, coverage centrality, and discrimination centrality.…”

Section: B Graph-based Approachesmentioning

confidence: 99%

“…The authors in [26] used an information extraction method based on BM25 and TF-IDF [27] to extract the candidate terms. Authors in [25] rely on graph-based methods for ranking the candidate terms. Authors in [29] consider the categories of extracted Wikipedia documents as candidates.…”

Section: ) Candidate Termsmentioning

confidence: 99%

Advanced Hierarchical Topic Labeling for Short Text

et al. 2023

View full text Add to dashboard Cite

Hierarchical Topic Modeling is the probabilistic approach for discovering latent topics distributed hierarchically among the documents. The distributed topics are represented with the respective topic terms. An unambiguous conclusion from the topic term distribution is a challenge for readers. The hierarchical topic labeling eases the challenge by facilitating an individual, appropriate label for each topic at every level. In this work, we propose a BERT-embedding inspired methodology for labeling hierarchical topics in short text corpora. The short texts have gained significant popularity on multiple platforms in diverse domains. The limited information available in the short text makes it difficult to deal with. In our work, we have used three diverse short text datasets that include both structured and unstructured instances. Such diversity ensures the broad application scope of this work. Considering the relevancy factor of the labels, the proposed methodology has been compared against both automatic and human annotators. Our proposed methodology outperformed the benchmark with an average score of 0.4185, 49.50, and 49.16 for cosine similarity, exact match, and partial match, respectively.

show abstract

“…For micro-blogging such as Twitter, text summarization using Text Rank [4] as well as hashtags [38] was also used successfully for topic labeling. Other techniques use word vectors and letter trigram vectors to automatically extract labels for topics [22] as well as unsupervised graph-based methods [1], [19], [44]. In the literature, there have also been proposed hybrid methods that make use of multiple measures to increase the chance to find the correct label for the topic [18].…”

Section: Related Workmentioning

confidence: 99%

TLATR: Automatic Topic Labeling Using Automatic (Domain-Specific) Term Recognition

Truică

Apostol

2021

IEEE Access

View full text Add to dashboard Cite

Topic modeling is a probabilistic graphical model for discovering latent topics in text corpora by using multinomial distributions of topics over words. Topic labeling is used to assign meaningful labels for the discovered topics. In this paper, we present a new topic labeling method that uses automatic term recognition to discover and assign relevant labels for each topic, i.e., TLATR (Topic Labeling using Automatic Term Recognition). TLATR uses domain-specific multi-terms that appear in the set of documents belonging to a topic. The multi-term having the highest score as determined by the automatic term recognition algorithm is chosen as the label for that topic. To evaluate TLATR, we use two real, publicly available datasets that contain scientific articles' abstracts. The topic label evaluation is done both automatically and using human annotators. For the automatic evaluation, we use Pointwise Mutual Information, Normalized Pointwise Mutual Information, and document similarity. For human evaluation, we employ the average rating method. Furthermore, we also evaluate the quality of the topic models using the Adjusted Rand Index. To prove that our novel method extracts relevant topic labels, we compare TLATR with two state-of-the-art methods, one supervised and one unsupervised, provided by the NETL Automatic Topic Labelling system. The experimental results show that our method outperforms or provides similar results with both NETL's supervised and unsupervised approaches.

show abstract

Automatic Labeling of Topic Models Using Graph-Based Ranking

Cited by 12 publications

References 37 publications

Musawah: A Data-Driven AI Approach and Tool to Co-Create Healthcare Services with a Case Study on Cancer Disease in Saudi Arabia

Musawah: A Data-Driven AI Approach and Tool to Co-Create Healthcare Services with a Case Study on Cancer Disease in Saudi Arabia

Advanced Hierarchical Topic Labeling for Short Text

TLATR: Automatic Topic Labeling Using Automatic (Domain-Specific) Term Recognition

Contact Info

Product

Resources

About