A phrase mining framework for recursive construction of a topical hierarchy

Wang, Chi; Danilevsky, Marina; Desai, Nihit; Zhang, Yinan; Nguyen, Phuong Anh; Taula, Thrivikrama; Han, Jiawei

doi:10.1145/2487575.2487631

Cited by 77 publications

(64 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…With respect to input and output, our definition of the construction of topical hierarchy largely follows our previous work Wang et al [2]. We proposed CATHY, a statisticsbased technique which constructs a topical hierarchy without resorting to external knowledge resources such as WordNet or Wikipedia.…”

Section: A Topical Hierarchy Constructionmentioning

confidence: 99%

“…(2). We use an Expectation-Maximization (EM) algorithm that can iteratively infer the model parameters.…”

Section: The Generative Modelmentioning

confidence: 99%

“…There are four criteria for judging the quality of a pattern (similar criteria are proposed for ranking phrases in [2], and also apply to other types of patterns).…”

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

“…Let N t be the number of documents that contain at least one frequent topic-t pattern, T a subset of C P ar(t) that contains t, and N T the number of documents that contain at least one frequent topic-z pattern for some topic z ∈ T . We use the following ranking function that satisfies all these requirements [2]:…”

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

“…The weight between exclusiveness and cohesiveness is controled by a parameter ω ∈ [0, +∞), with larger values of ω biasing the ranking more heavily towards cohesiveness. Due to space limitation, we refer to [2] for more detailed discussion of this ranking function.…”

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

See 4 more Smart Citations

Constructing topical hierarchies in heterogeneous information networks

et al. 2014

Self Cite

View full text Add to dashboard Cite

Abstract-A digital data collection (e.g., scientific publications, enterprise reports, news, and social media) can often be modeled as a heterogeneous information network, linking text with multiple types of entities. Constructing high-quality concept hierarchies that can represent topics at multiple granularities benefits tasks such as search, information browsing, and pattern mining. In this work we present an algorithm for recursively constructing multi-typed topical hierarchies. Contrary to traditional text-based topic modeling, our approach handles both textual phrases and multiple types of entities by a newly designed clustering and ranking algorithm for heterogeneous network data, as well as mining and ranking topical patterns of different types. Our experiments on datasets from two different domains demonstrate that our algorithm yields high quality, multi-typed topical hierarchies.

show abstract

Section: A Topical Hierarchy Constructionmentioning

confidence: 99%

“…(2). We use an Expectation-Maximization (EM) algorithm that can iteratively infer the model parameters.…”

Section: The Generative Modelmentioning

confidence: 99%

“…There are four criteria for judging the quality of a pattern (similar criteria are proposed for ranking phrases in [2], and also apply to other types of patterns).…”

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

Section: Pattern Ranking In Each Topicmentioning

confidence: 99%

See 3 more Smart Citations

Constructing topical hierarchies in heterogeneous information networks

et al. 2014

Self Cite

View full text Add to dashboard Cite

show abstract

Improving taxonomic relation learning via incorporating relation descriptions into word embeddings

Huang

Luo

Huang

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

SummaryTaxonomic relations play an important role in various Natural Language Processing (NLP) tasks (eg, information extraction, question answering and knowledge inference). Existing approaches on embedding‐based taxonomic relation learning mainly rely on the word embeddings trained using co‐occurrence‐based similarity learning. However, the performance of these approaches is not quite satisfactory due to the lack of sufficient taxonomic semantic knowledge within word embeddings. To solve this problem, we propose an improved embedding‐based approach to learn taxonomic relations via incorporating relation descriptions into word embeddings. First, to capture additional taxonomic semantic knowledge, we train special word embeddings using not only co‐occurrence information of words but also relation descriptions (eg, taxonomic seed relations and their contextual triples). Then, using the trained word embeddings as features, we employ two learning models to identify and predict taxonomic relations, namely, offset‐based classification model and offset‐based similarity model. Experimental results on four real‐world domain datasets demonstrate that our proposed approach can capture additional taxonomic semantic knowledge and reduce dependence on the training dataset, outperforming the state‐of‐the‐art compared approaches on the taxonomic relation learning task.

show abstract

Learning Concept Hierarchy from Short Texts Using Context Coherence

Almars

Ibrahim

et al. 2018

Web Information Systems Engineering – WISE 2018

View full text Add to dashboard Cite

A phrase mining framework for recursive construction of a topical hierarchy

Cited by 77 publications

References 21 publications

Constructing topical hierarchies in heterogeneous information networks

Constructing topical hierarchies in heterogeneous information networks

Improving taxonomic relation learning via incorporating relation descriptions into word embeddings

Learning Concept Hierarchy from Short Texts Using Context Coherence

Contact Info

Product

Resources

About