Fuzzy sets across the natural language generation pipeline

Ramos-Soto, Alejandro; Bugarín, Alberto; Barro, Senén

doi:10.1007/s13748-016-0097-x

Cited by 17 publications

(9 citation statements)

References 45 publications

(81 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have seen that our algorithms for the generation of quantified descriptions bear some striking resemblances with existing algorithms designed for a very different problem, namely REG. We plan to look at some recent advances of REG algorithms, such as those of Ramos-Soto et al (2016); Monroe and Potts (2015); Li et al (2018);van Gompel et al (2019) to see how they can inspire improvements of our QDG models.…”

Section: Discussion and Future Workmentioning

confidence: 99%

“…Quantified Expression Generation. Previous computational models of speakers' choice of quantifiers focus on the choice between a limited number of candidate quantifiers (Grefenstette, 2013;Yildirim et al, 2013;Herbelot and Vecchi, 2015;Sorodoc et al, 2016;Castillo-Ortega et al, 2009;Barr et al, 2013;Ramos-Soto et al, 2016). Barr et al (2013) studies the use of quantifiers that Most of the items are red circles, but there are a couple of blue squares.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Generating Quantified Descriptions of Abstract Visual Scenes

Chen

Deemter

Lin

2019

Proceedings of the 12th International Conference on Natural Language Generation

View full text Add to dashboard Cite

Quantified expressions have always taken up a central position in formal theories of meaning and language use. Yet quantified expressions have so far attracted far less attention from the Natural Language Generation community than, for example, referring expressions. In an attempt to start redressing the balance, we investigate a recently developed corpus in which quantified expressions play a crucial role; the corpus is the result of a carefully controlled elicitation experiment, in which human participants were asked to describe visually presented scenes. Informed by an analysis of this corpus, we propose algorithms that produce computer-generated descriptions of a wider class of visual scenes, and we evaluate the descriptions generated by these algorithms in terms of their correctness, completeness, and human-likeness. We discuss what this exercise can teach us about the nature of quantification and about the challenges posed by the generation of quantified expressions.

show abstract

Section: Discussion and Future Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Generating Quantified Descriptions of Abstract Visual Scenes

Chen

Deemter

Lin

2019

Proceedings of the 12th International Conference on Natural Language Generation

View full text Add to dashboard Cite

show abstract

“…Likewise, as we mentioned in Section 1, fuzzy sets have also been proposed in a more general way to be used in D2T systems to model vague terms [2,18,19]. To our knowledge, as of today the only D2T system which uses this kind of techniques and has been deployed in a real environment is GALiWeather [5].…”

Section: Related Workmentioning

confidence: 99%

“…Quite often, temporal and geographical expressions that need to be included in texts generated by D2T systems are vague, such as "in the evening" [4] or "southwestern areas" [17]. In situations where vagueness (and thus borderline cases and gradual concepts) is present, fuzzy sets have been proposed as a tool that allows to model linguistic expressions for NLG/D2T systems [2,18,19]. However, current existing D2T systems do not make use of such techniques, with the exception of GALiWeather [5], which provides a basic use of fuzzy sets to model temporal expressions and quantifiers.…”

Section: Introductionmentioning

confidence: 99%

Fuzzy-Based Language Grounding of Geographical References: From Writers to Readers

Ramos¹,

Alonso²,

Reiter³

et al. 2019

IJCIS

View full text Add to dashboard Cite

We describe an applied methodology to build fuzzy models of geographical expressions, which are meant to be used for natural language generation purposes. Our approach encompasses a language grounding task within the development of an actual datato-text system for the generation of textual descriptions of live weather data. For this, we gathered data from meteorologists through a survey and built consistent fuzzy models that aggregate the interpersonal variations found among the experts. A subset of the models was utilized in an illustrative use case, where we generated linguistic descriptions of weather maps for specific geographical expressions. These were used in a task-based evaluation to determine how well potential readers are able to identify the geographical expressions grounded on the models.

show abstract

“…The field of linguistic description of data (LDD) has emerged from the use of fuzzy set theory and soft computing to perform linguistic computations on data, which studies the methods for automatically describing numeric data sets by employing a set of linguistic terms (Ramos-Soto, Bugarín, & Barro, 2016). Fuzzy set theory is a well-studied approach to bridge between numeric and linguistic information, specifically in perception-based systems (Batyrshin et al, 2007;Kacprzyk & Zadrożny, 2010).…”

Section: Related Work On Semantic Inferences and Linguistic Descriptionsmentioning

confidence: 99%

Data-driven Conceptual Spaces: Creating Semantic Representations For Linguistic Descriptions Of Numerical Data

Banaee¹,

Schaffernicht²,

Loutfi³

2018

jair

View full text Add to dashboard Cite

There is an increasing need to derive semantics from real-world observations to facilitate natural information sharing between machine and human. Conceptual spaces theory is a possible approach and has been proposed as mid-level representation between symbolic and sub-symbolic representations, whereby concepts are represented in a geometrical space that is characterised by a number of quality dimensions. Currently, much of the work has demonstrated how conceptual spaces are created in a knowledge-driven manner, relying on prior knowledge to form concepts and identify quality dimensions. This paper presents a method to create semantic representations using data-driven conceptual spaces which are then used to derive linguistic descriptions of numerical data. Our contribution is a principled approach to automatically construct a conceptual space from a set of known observations wherein the quality dimensions and domains are not known a priori. This novelty of the approach is the ability to select and group semantic features to discriminate between concepts in a data-driven manner while preserving the semantic interpretation that is needed to infer linguistic descriptions for interaction with humans. Two data sets representing leaf images and time series signals are used to evaluate the method. An empirical evaluation for each case study assesses how well linguistic descriptions generated from the conceptual spaces identify unknown observations. Furthermore, comparisons are made with descriptions derived on alternative approaches for generating semantic models.

show abstract

Fuzzy sets across the natural language generation pipeline

Cited by 17 publications

References 45 publications

Generating Quantified Descriptions of Abstract Visual Scenes

Generating Quantified Descriptions of Abstract Visual Scenes

Fuzzy-Based Language Grounding of Geographical References: From Writers to Readers

Data-driven Conceptual Spaces: Creating Semantic Representations For Linguistic Descriptions Of Numerical Data

Contact Info

Product

Resources

About