On generating linguistic descriptions of time series

Processing Binary and Fuzzy Logic by Chaotic Time Series Generated by a Hydrodynamic Photochemical Oscillator

“…In fact, FLS(A) is less accurate than FLS(B) in describing the training data because it is based on a smaller number of fuzzy sets; nevertheless, it is more meaningful in predicting the testing data. This observation is relevant for the application of these fuzzy logic systems in the field of computational linguistics . In fact, if we transform the two optimized Sugeno FLSs (A) and (B) into two Mamdani FLSs and label each fuzzy set by adjectives, we can exploit them to describe the aperiodic time series generated by the MPh hydrodynamic photochemical oscillator through syllogistic statements involving words of natural language.…”

“…Artificial neural networks, such as feed‐forward neural networks, are more powerful in this respect . However, the fuzzy logic system can be really useful in the field of computational linguistics, which is concerned with the development of systems suitable for generating texts in natural language from numerical data . Such data‐to‐text systems are playing a role in field of artificial intelligence and promise to be useful for making decisions in complex situations.…”

“…This case constitutes the fourth strategy to process fuzzy logic by using chemical systems, after the first, which was based on smooth analog linear or nonlinear macroscopic input–output relations,– the second, which was grounded on molecular conformers, and the third, which exploited the absorption bands of electronic spectra as fuzzy sets . This fourth strategy for implementing fuzzy logic systems has implications for the relatively recent research area of computational linguistics, which is concerned with the development of systems able to generate texts in natural language from numerical data . Such data‐to‐text systems are playing a relevant role in the field of artificial intelligence and can be useful for making decisions in complex situations.…”

The Role of Graduality for Referring Expression Generation in Visual Scenes

“…In order to asses the goodness or adequacy of a referring expression, other quality dimensions can be considered, apart from the ones previously discussed, similarly to what happens in complete linguistic descriptions [3]. The reg literature contains a number of such proposals.…”

“…In recent years, scene description has attracted a lot of attention as part of a growing trend to explore the interface between vision and language in Natural Language Generation (nlg) [1], as is the case of other kind of data [2][3][4].…”