Research Publications

2016

Hinkelmann K, Gerber A, Karagiannis D, Thoenssen B, van der Merwe A, Woitsch R. A new paradigm for the continuous alignment of business and IT: Combining enterprise architecture modelling and enterprise ontology. Computers in Industry. 2016;79. http://www.sciencedirect.com/science/article/pii/S0166361515300270.

No Abstract

@article{148,
  author = {Knut Hinkelmann and Aurona Gerber and Dimitris Karagiannis and Barbara Thoenssen and Alta van der Merwe and Robert Woitsch},
  title = {A new paradigm for the continuous alignment of business and IT: Combining enterprise architecture modelling and enterprise ontology},
  abstract = {No Abstract},
  year = {2016},
  journal = {Computers in Industry},
  volume = {79},
  publisher = {Sciencedirect},
  url = {http://www.sciencedirect.com/science/article/pii/S0166361515300270},
}
van der Merwe B. The Output Size Problem for String-to-Tree Transducers. 2016. http://www.ims.uni-stuttgart.de/events/TTATT2016.

No Abstract

@misc{146,
  author = {Brink van der Merwe},
  title = {The Output Size Problem for String-to-Tree Transducers},
  abstract = {No Abstract},
  year = {2016},
  url = {http://www.ims.uni-stuttgart.de/events/TTATT2016},
}
van der Merwe B. Analyzing Matching Time Behavior of Backtracking Regular Expression Matchers by Using Ambiguity of NFA. . 21st International Conference on Implementation and Application of Automata. 2016.

No Abstract

@proceedings{145,
  author = {Brink van der Merwe},
  title = {Analyzing Matching Time Behavior of Backtracking Regular Expression Matchers by Using Ambiguity of NFA.},
  abstract = {No Abstract},
  year = {2016},
  journal = {21st International Conference on Implementation and Application of Automata},
  pages = {322-334},
  month = {19/07-22/07},
}
Rens G, Meyer T, Casini G. On Revision of Partially Specified Convex Probabilistic Belief Bases. European Conference on Artificial Intelligence (ECAI). 2016.

We propose a method for an agent to revise its incomplete probabilistic beliefs when a new piece of propositional information is observed. In this work, an agent’s beliefs are represented by a set of probabilistic formulae – a belief base. The method involves determining a representative set of ‘boundary’ probability distributions consistent with the current belief base, revising each of these probability distributions and then translating the revised information into a new belief base. We use a version of Lewis Imaging as the revision operation. The correctness of the approach is proved. An analysis of the approach is done against six rationality postulates. The expressivity of the belief bases under consideration are rather restricted, but has some applications. We also discuss methods of belief base revision employing the notion of optimum entropy, and point out some of the benefits and difficulties in those methods. Both the boundary distribution method and the optimum entropy method are reasonable, yet yield different results.

@proceedings{144,
  author = {Gavin Rens and Thomas Meyer and Giovanni Casini},
  title = {On Revision of Partially Specified Convex Probabilistic Belief Bases},
  abstract = {We propose a method for an agent to revise its incomplete probabilistic beliefs when a new piece of propositional information is observed. In this work, an agent’s beliefs are represented by a set of probabilistic formulae – a belief base. The method involves determining a representative set of ‘boundary’ probability distributions consistent with the current belief base, revising each of these probability distributions and then translating the revised information into a new belief base. We use a version of Lewis Imaging as the revision operation. The correctness of the approach is proved. An analysis of the approach is done against six rationality postulates. The expressivity of the belief bases under consideration are rather restricted, but has some applications. We also discuss methods of belief base revision employing the notion of optimum entropy, and point out some of the benefits and difficulties in those methods. Both the boundary distribution method and the optimum entropy method are reasonable, yet yield different results.},
  year = {2016},
  journal = {European Conference on Artificial Intelligence (ECAI)},
  pages = {921-929},
  month = {31/08-02/09},
}
Kroon S, Le Roux PB, Bester W. DSaaS: A cloud Service for Persistent Data Structures. CLOSER, 6th International Conference on Cloud Computing and Services Science. 2016.

No Abstract

@proceedings{143,
  author = {Steve Kroon and PB Le Roux and Willem Bester},
  title = {DSaaS:  A cloud Service for Persistent Data Structures},
  abstract = {No Abstract},
  year = {2016},
  journal = {CLOSER, 6th International Conference on Cloud Computing and Services Science},
  pages = {37-48},
  month = {23/04-25/04},
  address = {Portugal},
  isbn = {978-989-758-182-3},
}
de Villiers H, Wiehman S. Semantic Segmentation of Bioimages Using Convolutional Neural Networks. 2016 International Joint Conference on Neural Networks (IJCNN). 2016.

Convolutional neural networks have shown great promise in both general image segmentation problems as well as bioimage segmentation. In this paper, the application of different convolutional network architectures is explored on the C. elegans live/dead assay dataset from the Broad Bioimage Benchmark Collection. These architectures include a standard convolutional network which produces single pixel outputs, as well as Fully Convolutional Networks (FCN) for patch prediction. It was shown that the custom image processing pipeline, which achieved a worm segmentation accuracy of 94%, was outperformed by all of the architectures considered, with the best being 97.3% achieved by a FCN with a single downsampling layer. These results demonstrate the promise of employing convolutional neural network architectures as an alternative to ad-hoc image processing pipelines on optical microscopy images of C. elegans.

@proceedings{142,
  author = {Hennie de Villiers and Stiaan Wiehman},
  title = {Semantic Segmentation of Bioimages Using Convolutional Neural Networks},
  abstract = {Convolutional neural networks have shown great promise in both general image segmentation problems as well as bioimage segmentation. In this paper, the application of different convolutional network architectures is explored on the C. elegans live/dead assay dataset from the Broad Bioimage Benchmark Collection. These architectures include a standard convolutional network which produces single pixel outputs, as well as Fully Convolutional Networks (FCN) for patch prediction. It was shown that the custom image processing pipeline, which achieved a worm segmentation accuracy of 94%, was outperformed by all of the architectures considered, with the best being 97.3% achieved by a FCN with a single downsampling layer. These results demonstrate the promise of employing convolutional neural network architectures as an alternative to ad-hoc image processing pipelines on optical microscopy images of C. elegans.},
  year = {2016},
  journal = {2016 International Joint Conference on Neural Networks (IJCNN)},
  pages = {624-631},
  month = {24/07-29/07},
  address = {Singapore},
  isbn = {978-1-5090-0620-5},
}
Greene GJ, Fischer B. CVExplorer: Identifying Candidate Developers by Mining and Exploring Their Open Source Contributions. Automated Software Engineering. 2016.

No Abstract

@proceedings{141,
  author = {G.J. Greene and Bernd Fischer},
  title = {CVExplorer: Identifying Candidate Developers by Mining and Exploring Their Open Source Contributions},
  abstract = {No Abstract},
  year = {2016},
  journal = {Automated Software Engineering},
  pages = {804-809},
  month = {03/09-07/09},
  isbn = {978-1-4503-3845-5},
}
Rens G. A Stochastic Belief Change Framework with an Observation Stream and Defaults as Expired Observations. 2016.

A framework for an agent to change its probabilistic beliefs after a stream of noisy observations is received is proposed. Observations which are no longer relevant, become default assumptions until overridden by newer, more prevalent observations. A distinction is made between background and foreground beliefs. Agent actions and environment events are distinguishable and form part of the agent model. It is left up to the agent designer to provide an environment model; a submodel of the agent model. An example of an environment model is provided in the paper, and an example scenario is based on it. Given the particular form of the agent model, several ‘patterns of cognition’ can be identified. An argument is made for four particular patterns.

@misc{140,
  author = {Gavin Rens},
  title = {A Stochastic Belief Change Framework with an Observation Stream and Defaults as Expired Observations},
  abstract = {A framework for an agent to change its probabilistic beliefs after a stream of noisy observations is received is proposed. Observations which are no longer relevant, become default assumptions until overridden by newer, more prevalent observations. A distinction is made between background and foreground beliefs. Agent actions and environment events are distinguishable and form part of the agent model. It is left up to the agent designer to provide an environment model; a submodel of the agent model. An example of an environment model is provided in the paper, and an example scenario is based on it. Given the particular form of the agent model, several ‘patterns of cognition’ can be identified. An argument is made for four particular patterns.},
  year = {2016},
}
Rens G, Kern-Isberner G. An Approach to Qualitative Belief Change Modulo Ontic Strength. 2016.

Sometimes, strictly choosing between belief revision and belief update is inadequate in a dynamical, uncertain environment. Boutilier combined the two notions to allow updates in response to external changes to inform an agent about its prior beliefs. His approach is based on ranking functions. Rens proposed a new method to trade off probabilistic revision and update, in proportion to the agent’s confidence for whether to revise or update. In this paper, we translate Rens’s approach from a probabilistic setting to a setting with ranking functions. Given the translation, we are able to compare Boutilier’s and Rens’s approaches.We found that Rens’s approach is an extension of Boutilier’s.

@misc{139,
  author = {Gavin Rens and Gabriele Kern-Isberner},
  title = {An Approach to Qualitative Belief Change Modulo Ontic Strength},
  abstract = {Sometimes, strictly choosing between belief revision and belief update is inadequate in a dynamical, uncertain environment. Boutilier combined the two notions to allow updates in response to external changes to inform an agent about its prior beliefs. His approach is based on ranking functions. Rens proposed a new method to trade off probabilistic revision and update, in proportion to the agent’s confidence for whether to revise or update. In this paper, we translate Rens’s approach from a probabilistic setting to a setting with ranking functions. Given the translation, we are able to compare Boutilier’s and Rens’s approaches.We found that Rens’s approach is an extension of Boutilier’s.},
  year = {2016},
}
Greene GJ, Fischer B. Single-Focus Broadening Navigation in Concept Lattices. 2016. https://cla2016.hse.ru/data/2016/07/24/1119022942/CDUD2016.pdf#page=39.

No Abstract

@misc{138,
  author = {G.J. Greene and Bernd Fischer},
  title = {Single-Focus Broadening Navigation in Concept Lattices},
  abstract = {No Abstract},
  year = {2016},
  url = {https://cla2016.hse.ru/data/2016/07/24/1119022942/CDUD2016.pdf#page=39},
}
Britz K, Varzinczak I. Preferential modalities revisited. 2016. http://nmr2016.cs.uct.ac.za/proceedings_nmr2016_online.pdf.

We venture beyond the customary semantic approach in NMR, namely that of placing orderings on worlds (or valuations). In a modal-logic setting, we motivate and investigate the idea of ordering elements of the accessibility relations in Kripke frames, i.e, world pairs (w,w') (or arrows). The underlying intuition is that some world pairs may be seen as more normal (or typical, or expected) than others. We show this delivers an elegant and intuitive semantic construction, which gives a new perspective on present notions of defeasible necessity. From a modeler's perspective, the new framework we propose is more intuitively appealing. Technically, though, the revisited logic happens to not substantively increase the expressive power of the previously defined preferential modalities. This conclusion follows from an analysis of both semantic constructions via a generalisation of bisimulations to the preferential case. Lest this be seen as a negative result, it essentially means that reasoners based on the previous semantics (which have been shown to preserve the computational complexity of the underlying classical modal language) suffice for reasoning over the new semantics. Finally, we show that the kind of construction we here propose has many fruitful applications, notably in a description-logic context, where it provides the foundations on which to ground useful notions of defeasibility in ontologies yet to be explored.

@misc{135,
  author = {Katarina Britz and Ivan Varzinczak},
  title = {Preferential modalities revisited},
  abstract = {We venture beyond the customary semantic approach in NMR, namely that of placing orderings on worlds (or valuations). In a modal-logic setting, we motivate and investigate the idea of ordering elements of the accessibility relations in Kripke frames, i.e, world pairs (w,w') (or arrows). The underlying intuition is that some world pairs may be seen as more normal (or typical, or expected) than others. We show this delivers an elegant and intuitive semantic construction, which gives a new perspective on present notions of defeasible necessity. From a modeler's perspective, the new framework we propose is more intuitively appealing. Technically, though, the revisited logic happens to not substantively increase the expressive power of the previously defined preferential modalities. This conclusion follows from an analysis of both semantic constructions via a generalisation of bisimulations to the preferential case. Lest this be seen as a negative result, it essentially means that reasoners based on the previous semantics (which have been shown to preserve the computational complexity of the underlying classical modal language) suffice for reasoning over the new semantics. Finally, we show that the kind of construction we here propose has many fruitful applications, notably in a description-logic context, where it provides the foundations on which to ground useful notions of defeasibility in ontologies yet to be explored.},
  year = {2016},
  url = {http://nmr2016.cs.uct.ac.za/proceedings_nmr2016_online.pdf},
}
Rens G. On Stochastic Belief Revision and Update and their Combination. 2016.

I propose a framework for an agent to change its probabilistic beliefs when a new piece of propositional information alpha is observed. Traditionally, belief change occurs by either a revision process or by an update process, depending on whether the agent is informed with alpha in a static world or, respectively, whether alpha is a ‘signal’ from the environment due to an event occurring. Boutilier suggested a unified model of qualitative belief change, which “combines aspects of revision and update, providing a more realistic characterization of belief change.” In this paper, I propose a unified model of quantitative belief change, where an agent’s beliefs are represented as a probability distribution over possible worlds. As does Boutilier, I take a dynamical systems perspective. The proposed approach is evaluated against several rationality postulated, and some properties of the approach are worked out.

@misc{132,
  author = {Gavin Rens},
  title = {On Stochastic Belief Revision and Update and their Combination},
  abstract = {I propose a framework for an agent to change its probabilistic beliefs when a new piece of propositional information alpha is observed. Traditionally, belief change occurs by either a revision process or by an update process, depending on whether the agent is informed with alpha in a static world or, respectively, whether alpha is a ‘signal’ from the environment due to an event occurring. Boutilier suggested a unified model of qualitative belief change, which “combines aspects of revision and update, providing a more realistic characterization of belief change.” In this paper, I propose a unified model of quantitative belief change, where an agent’s beliefs are represented as a probability distribution over possible worlds. As does Boutilier, I take a dynamical systems perspective. The proposed approach is evaluated against several rationality postulated, and some properties of the approach are worked out.},
  year = {2016},
  isbn = {ISSN 0933-6192},
}
Rens G, Meyer T, Casini G. Revising Incompletely Specified Convex Probabilistic Belief Bases. 2016.

We propose a method for an agent to revise its incomplete probabilistic beliefs when a new piece of propositional information is observed. In this work, an agent’s beliefs are represented by a set of probabilistic formulae – a belief base. The method involves determining a representative set of ‘boundary’ probability distributions consistent with the current belief base, revising each of these probability distributions and then translating the revised information into a new belief base. We use a version of Lewis Imaging as the revision operation. The correctness of the approach is proved. The expressivity of the belief bases under consideration are rather restricted, but has some applications. We also discuss methods of belief base revision employing the notion of optimum entropy, and point out some of the benefits and difficulties in those methods. Both the boundary distribution method and the optimum entropy method are reasonable, yet yield different results.

@misc{131,
  author = {Gavin Rens and Thomas Meyer and Giovanni Casini},
  title = {Revising Incompletely Specified Convex Probabilistic Belief Bases},
  abstract = {We propose a method for an agent to revise its incomplete probabilistic beliefs when a new piece of propositional information is observed. In this work, an agent’s beliefs are represented by a set of probabilistic formulae – a belief base. The method involves determining a representative set of ‘boundary’ probability distributions consistent with the current belief base, revising each of these probability distributions and then translating the revised information into a new belief base. We use a version of Lewis Imaging as the revision operation. The correctness of the approach is proved. The expressivity of the belief bases under consideration are rather restricted, but has some applications. We also discuss methods of belief base revision employing the notion of optimum entropy, and point out some of the benefits and difficulties in those methods. Both the boundary distribution method and the optimum entropy method are reasonable, yet yield different results.},
  year = {2016},
  isbn = {ISSN 0933-6192},
}
Kleynhans N, Hartman W, van Niekerk D, et al. Code-switched English Pronunciation Modeling for Swahili Spoken Term Detection. Workshop on Spoken Language Technologies for Under-Resourced Languages. 2016.

No Abstract

@proceedings{133,
  author = {N. Kleynhans and W. Hartman and D. van Niekerk and Charl van Heerden and R. Schwartz and S. Tsakalidis and Marelie Davel},
  title = {Code-switched English Pronunciation Modeling for Swahili Spoken Term Detection},
  abstract = {No Abstract},
  year = {2016},
  journal = {Workshop on Spoken Language Technologies for Under-Resourced Languages},
  pages = {128-135},
  month = {09/05-12/05},
  address = {online},
  isbn = {1877-0509},
}
Waltham M, Moodley D. An Analysis of Artificial Intelligence Techniques in Multiplayer Online Battle Arena Game Environments. ACM Proceedings of the Annual Conference of the South African Institute of Computer Scientists and Information Technologists. 2016. doi: http://dx.doi.org/10.1145/2987491.2987513.

The 3D computer gaming industry is constantly exploring new avenues for creating immersive and engaging environments. One avenue being explored is autonomous control of the behaviour of non-player characters (NPC). This paper reviews and compares existing artificial intelligence (AI) techniques for controlling the behaviour of non-human characters in Multiplayer Online Battle Arena (MOBA) game environments. Two techniques, the fuzzy state machine (FuSM) and the emotional behaviour tree (EBT),were reviewed and compared. In addition, an alternate and simple mechanism to incorporate emotion in a behaviour tree is proposed and tested. Initial tests of the mechanism show that it is a viable and promising mechanism for effectively tracking the emotional state of an NPC and for incorporating emotion in NPC decision making.

@proceedings{251,
  author = {Michael Waltham and Deshen Moodley},
  title = {An Analysis of Artificial Intelligence Techniques in Multiplayer Online Battle Arena Game Environments},
  abstract = {The 3D computer gaming industry is constantly exploring new avenues for creating immersive and engaging environments. One avenue being explored is autonomous control of the behaviour of non-player characters (NPC). This paper reviews and compares existing artificial intelligence (AI) techniques for controlling the behaviour of non-human characters in Multiplayer Online Battle Arena (MOBA) game environments. Two techniques, the fuzzy state machine (FuSM) and the emotional behaviour tree (EBT),were reviewed and compared. In addition, an alternate and simple mechanism to incorporate emotion in a behaviour tree is proposed and tested. Initial tests of the mechanism show that it is a viable and promising mechanism for effectively tracking the emotional state of an NPC and for incorporating emotion in NPC decision making.},
  year = {2016},
  journal = {ACM Proceedings of the Annual Conference of the South African Institute of Computer Scientists and Information Technologists},
  pages = {45},
  month = {26/09 - 28/09},
  publisher = {ACM},
  address = {Johannesburg},
  isbn = {978-1-4503-4805-8/16/09},
  doi = {http://dx.doi.org/10.1145/2987491.2987513},
}

2015

Booth R, Casini G, Meyer T, Varzinczak I. On the Entailment Problem for a Logic of Typicality. Twenty-Fourth International Joint Conference on Artificial Intelligence (IJCAI). 2015. https://www.semanticscholar.org/paper/On-the-Entailment-Problem-for-a-Logic-of-Typicality-Booth-Casini/47bb88f18d591df1355c7ad53593a1f1b46522f0.

Propositional Typicality Logic (PTL) is a recently proposed logic, obtained by enriching classical propositional logic with a typicality operator. In spite of the non-monotonic features introduced by the semantics adopted for the typicality operator, the obvious Tarskian definition of entailment for PTL remains monotonic and is therefore not appropriate. We investigate different (semantic) versions of entailment for PTL, based on the notion of Rational Closure as defined by Lehmann and Magidor for KLM-style conditionals, and constructed using minimality. Our first important result is an impossibility theorem showing that a set of proposed postulates that at first all seem appropriate for a notion of entailment with regard to typicality cannot be satisfied simultaneously. Closer inspection reveals that this result is best interpreted as an argument for advocating the development of more than one type of PTL entailment. In the spirit of this interpretation, we define two primary forms of entailment for PTL and discuss their advantages and disadvantages

@proceedings{363,
  author = {Richard Booth and Giovanni Casini and Thomas Meyer and Ivan Varzinczak},
  title = {On the Entailment Problem for a Logic of Typicality},
  abstract = {Propositional Typicality Logic (PTL) is a recently proposed logic, obtained by enriching classical propositional logic with a typicality operator. In spite of the non-monotonic features introduced by the semantics adopted for the typicality operator, the obvious Tarskian definition of entailment for PTL remains monotonic and is therefore not appropriate. We investigate different (semantic) versions of entailment for PTL, based on the notion of Rational Closure as defined by Lehmann and Magidor for KLM-style conditionals, and constructed using minimality. Our first important result is an impossibility theorem showing that a set of proposed postulates that at first all seem appropriate for a notion of entailment with regard to typicality cannot be satisfied simultaneously. Closer inspection reveals that this result is best interpreted as an argument for advocating the development of more than one type of PTL entailment. In the spirit of this interpretation, we define two primary forms of entailment for PTL and discuss their advantages and disadvantages},
  year = {2015},
  journal = {Twenty-Fourth International Joint Conference on Artificial Intelligence (IJCAI)},
  pages = {2805-2811},
  month = {25/07 - 31/07},
  publisher = {AAAI Press},
  url = {https://www.semanticscholar.org/paper/On-the-Entailment-Problem-for-a-Logic-of-Typicality-Booth-Casini/47bb88f18d591df1355c7ad53593a1f1b46522f0},
}
Rens G, Meyer T. A New Approach to Probabilistic Belief Change. International Florida AI Research Society Conference (FLAIRS). 2015. https://www.researchgate.net/publication/277004602_A_New_Approach_to_Probabilistic_Belief_Change.

One way for an agent to deal with uncertainty about its beliefs is to maintain a probability distribution over the worlds it believes are possible. A belief change operation may recommend some previously believed worlds to become impossible and some previously disbelieved worlds to become possible. This work investigates how to redistribute probabilities due to worlds being added to and removed from an agent’s belief-state. Two related approaches are proposed and analyzed.

@proceedings{362,
  author = {Gavin Rens and Thomas Meyer},
  title = {A New Approach to Probabilistic Belief Change},
  abstract = {One way for an agent to deal with uncertainty about its beliefs is to maintain a probability distribution over the worlds it believes are possible. A belief change operation may recommend some previously believed worlds to become impossible and some previously disbelieved worlds to become possible. This work investigates how to redistribute probabilities due to worlds being added to and removed from an agent’s belief-state. Two related approaches are proposed and analyzed.},
  year = {2015},
  journal = {International Florida AI Research Society Conference (FLAIRS)},
  pages = {582-587},
  month = {18/05 - 20/05},
  publisher = {AAAI Press},
  url = {https://www.researchgate.net/publication/277004602_A_New_Approach_to_Probabilistic_Belief_Change},
}
Wissing D, Pienaar W, Van Niekerk DR. Palatalisation of /s/ in Afrikaans. Stellenbosch Papers in Linguistics Plus. 2015;48. doi:10.5842/48-0-688.

This article reports on the investigation of the acoustic characteristics of the Afrikaans voiceless alveolar fricative /s/[1]. As yet, a palatal [ʃ] for /s/ has been reported only in a limited case, namely where /s/ is followed by palatal /j/, for example in the phrase is jy (‘are you’), pronounced as [ə-ʃəi]. This seems to be an instance of regressive coarticulation, resulting in coalescence of basic /s/ and /j/. The present study revealed that, especially in the pronunciation of young, white Afrikaans-speakers, /s/ is also palatalised progressively when preceded by /r/ in the coda cluster /rs/, and, to a lesser extent, also in other contexts where /r/ is involved, for example across syllable and word boundaries. Only a slight presence of palatalisation was detected in the production of /s/ in the speech of the white, older speakers of the present study. This finding might be indicative of a definite change in the Afrikaans consonant system. A post hoc reflection is offered here on the possible presence of /s/-fronting, especially in the speech of the younger females. Such pronunciation could very well be a prestige marker for affluent speakers of Afrikaans.

@article{293,
  author = {Daan Wissing and Wikus Pienaar and Daniel Van Niekerk},
  title = {Palatalisation of /s/ in Afrikaans},
  abstract = {This article reports on the investigation of the acoustic characteristics of the Afrikaans voiceless alveolar fricative /s/[1]. As yet, a palatal [ʃ] for /s/ has been reported only in a limited case, namely where /s/ is followed by palatal /j/, for example in the phrase is jy (‘are you’), pronounced as [ə-ʃəi]. This seems to be an instance of regressive coarticulation, resulting in coalescence of basic /s/ and /j/. The present study revealed that, especially in the pronunciation of young, white Afrikaans-speakers, /s/ is also palatalised progressively when preceded by /r/ in the coda cluster /rs/, and, to a lesser extent, also in other contexts where /r/ is involved, for example across syllable and word boundaries. Only a slight presence of palatalisation was detected in the production of /s/ in the speech of the white, older speakers of the present study. This finding might be indicative of a definite change in the Afrikaans consonant system. A post hoc reflection is offered here on the possible presence of /s/-fronting, especially in the speech of the younger females. Such pronunciation could very well be a prestige marker for affluent speakers of Afrikaans.},
  year = {2015},
  journal = {Stellenbosch Papers in Linguistics Plus},
  volume = {48},
  pages = {137-158},
  publisher = {Stellenbosch University},
  doi = {10.5842/48-0-688},
}
Modipa T, Davel MH. Predicting vowel substitution in code-switched speech. Pattern Recognition Association of South Africa (PRASA). 2015. doi:10.1109/RoboMech.2015.7359515.

The accuracy of automatic speech recognition (ASR) systems typically degrades when encountering code-switched speech. Some of this degradation is due to the unexpected pronunciation effects introduced when languages are mixed. Embedded (foreign) phonemes typically show more variation than phonemes from the matrix language: either approximating the embedded language pronunciation fairly closely, or realised as any of a set of phonemic counterparts from the matrix language. In this paper we describe a technique for predicting the phoneme substitutions that are expected to occur during code-switching, using non-acoustic features only. As case study we consider Sepedi/English code switching and analyse the different realisations of the English schwa. A code-switched speech corpus is used as input and vowel substitutions identified by auto-tagging this corpus based on acoustic characteristics. We first evaluate the accuracy of our auto-tagging process, before determining the predictability of our auto-tagged corpus, using non-acoustic features.

@proceedings{292,
  author = {Thipe Modipa and Marelie Davel},
  title = {Predicting vowel substitution in code-switched speech},
  abstract = {The accuracy of automatic speech recognition (ASR) systems typically degrades when encountering code-switched speech. Some of this degradation is due to the unexpected pronunciation effects introduced when languages are mixed. Embedded (foreign) phonemes typically show more variation than phonemes from the matrix language: either approximating the embedded language pronunciation fairly closely, or realised as any of a set of phonemic counterparts from the matrix language. In this paper we describe a technique for predicting the phoneme substitutions that are expected to occur during code-switching, using non-acoustic features only. As case study we consider Sepedi/English code switching and analyse the different realisations of the English schwa. A code-switched speech corpus is used as input and vowel substitutions identified by auto-tagging this corpus based on acoustic characteristics. We first evaluate the accuracy of our auto-tagging process, before determining the predictability of our auto-tagged corpus, using non-acoustic features.},
  year = {2015},
  journal = {Pattern Recognition Association of South Africa (PRASA)},
  pages = {154-159},
  month = {26/11-27/11},
  address = {Port Elizabeth, South Africa},
  isbn = {978-1-4673-7450-7, 978-1-4673-7449-1},
  doi = {10.1109/RoboMech.2015.7359515},
}
Kleynhans N, Barnard E. Efficient data selection for ASR. Language Resources and Evaluation. 2015;49(2). doi:10.1007/s10579-014-9285-0.

Automatic speech recognition (ASR) technology has matured over the past few decades and has made significant impacts in a variety of fields, from assistive technologies to commercial products. However, ASR system development is a resource intensive activity and requires language resources in the form of text annotated audio recordings and pronunciation dictionaries. Unfortunately, many languages found in the developing world fall into the resource-scarce category and due to this resource scarcity the deployment of ASR systems in the developing world is severely inhibited. One approach to assist with resource-scarce ASR system development, is to select 'useful' training samples which could reduce the resources needed to collect new corpora. In this work, we propose a new data selection framework which can be used to design a speech recognition corpus. We show for limited data sets, independent of language and bandwidth, the most effective strategy for data selection is frequency-matched selection and that the widely-used maximum entropy methods generally produced the least promising results. In our model, the frequency-matched selection method corresponds to a logarithmic relationship between accuracy and corpus size; we also investigated other model relationships, and found that a hyperbolic relationship (as suggested from simple asymptotic arguments in learning theory) may lead to somewhat better performance under certain conditions.

@article{291,
  author = {Neil Kleynhans and Etienne Barnard},
  title = {Efficient data selection for ASR},
  abstract = {Automatic speech recognition (ASR) technology has matured over the past few decades and has made significant impacts in a variety of fields, from assistive technologies to commercial products. However, ASR system development is a resource intensive activity and requires language resources in the form of text annotated audio recordings and pronunciation dictionaries. Unfortunately, many languages found in the developing world fall into the resource-scarce category and due to this resource scarcity the deployment of ASR systems in the developing world is severely inhibited. One approach to assist with resource-scarce ASR system development, is to select 'useful' training samples which could reduce the resources needed to collect new corpora. In this work, we propose a new data selection framework which can be used to design a speech recognition corpus. We show for limited data sets, independent of language and bandwidth, the most effective strategy for data selection is frequency-matched selection and that the widely-used maximum entropy methods generally produced the least promising results. In our model, the frequency-matched selection method corresponds to a logarithmic relationship between accuracy and corpus size; we also investigated other model relationships, and found that a hyperbolic relationship (as suggested from simple asymptotic arguments in learning theory) may lead to somewhat better performance under certain conditions.},
  year = {2015},
  journal = {Language Resources and Evaluation},
  volume = {49},
  pages = {327-353},
  issue = {2},
  publisher = {Springer Science+Business Media},
  address = {Dordrecht},
  doi = {10.1007/s10579-014-9285-0},
}
Kleynhans N, De Wet F, Barnard E. Unsupervised acoustic model training: comparing South African English and isiZulu. Pattern Recognition Association of South Africa (PRASA). 2015. doi: 10.1109/RoboMech.2015.7359512.

Large amounts of untranscribed audio data are generated every day. These audio resources can be used to develop robust acoustic models that can be used in a variety of speech-based systems. Manually transcribing this data is resource intensive and requires funding, time and expertise. Lightly-supervised training techniques, however, provide a means to rapidly transcribe audio, thus reducing the initial resource investment to begin the modelling process. Our findings suggest that the lightly-supervised training technique works well for English but when moving to an agglutinative language, such as isiZulu, the process fails to achieve the performance seen for English. Additionally, phone-based performances are significantly worse when compared to an approach using word-based language models. These results indicate a strong dependence on large or well-matched text resources for lightly-supervised training techniques.

@proceedings{290,
  author = {Neil Kleynhans and Febe De Wet and Etienne Barnard},
  title = {Unsupervised acoustic model training: comparing South African English and isiZulu},
  abstract = {Large amounts of untranscribed audio data are generated every day. These audio resources can be used to develop robust acoustic models that can be used in a variety of speech-based systems. Manually transcribing this data is resource intensive and requires funding, time and expertise. Lightly-supervised training techniques, however, provide a means to rapidly transcribe audio, thus reducing the initial resource investment to begin the modelling process. Our findings suggest that the lightly-supervised training technique works well for English but when moving to an agglutinative language, such as isiZulu, the process fails to achieve the performance seen for English. Additionally, phone-based performances are significantly worse when compared to an approach using word-based language models. These results indicate a strong dependence on large or well-matched text resources for lightly-supervised training techniques.},
  year = {2015},
  journal = {Pattern Recognition Association of South Africa (PRASA)},
  pages = {136-141},
  address = {Port Elizabeth, South Africa},
  isbn = {978-1-4673-7450-7, 978-1-4673-7449-1},
  doi = {10.1109/RoboMech.2015.7359512},
}
Giwa O, Davel MH. Text-based Language Identification of Multilingual Names. Pattern Recognition Association of South Africa (PRASA). 2015. doi: 10.1109/RoboMech.2015.7359517.

Text-based language identification (T-LID) of isolated words has been shown to be useful for various speech processing tasks, including pronunciation modelling and data categorisation. When the words to be categorised are proper names, the task becomes more difficult: not only do proper names often have idiosyncratic spellings, they are also often considered to be multilingual. We, therefore, investigate how an existing T-LID technique can be adapted to perform multilingual word classification. That is, given a proper name, which may be either mono- or multilingual, we aim to determine how accurately we can predict how many possible source languages the word has, and what they are. Using a Joint Sequence Model-based approach to T-LID and the SADE corpus - a newly developed proper names corpus of South African names - we experiment with different approaches to multilingual T-LID. We compare posterior-based and likelihood-based methods and obtain promising results on a challenging task.

@proceedings{289,
  author = {Oluwapelumi Giwa and Marelie Davel},
  title = {Text-based Language Identification of Multilingual Names},
  abstract = {Text-based language identification (T-LID) of isolated words has been shown to be useful for various speech processing tasks, including pronunciation modelling and data categorisation. When the words to be categorised are proper names, the task becomes more difficult: not only do proper names often have idiosyncratic spellings, they are also often considered to be multilingual. We, therefore, investigate how an existing T-LID technique can be adapted to perform multilingual word classification. That is, given a proper name, which may be either mono- or multilingual, we aim to determine how accurately we can predict how many possible source languages the word has, and what they are. Using a Joint Sequence Model-based approach to T-LID and the SADE corpus - a newly developed proper names corpus of South African names - we experiment with different approaches to multilingual T-LID. We compare posterior-based and likelihood-based methods and obtain promising results on a challenging task.},
  year = {2015},
  journal = {Pattern Recognition Association of South Africa (PRASA)},
  pages = {166-171},
  address = {Port Elizabeth, South Africa},
  isbn = {978-1-4673-7450-7, 978-1-4673-7449-1},
  doi = {10.1109/RoboMech.2015.7359517},
}
Davel MH, Barnard E, Van Heerden CJ, et al. Exploring minimal pronunciation modeling for low resource languages. Interspeech. 2015.

Pronunciation lexicons can range from fully graphemic (modeling each word using the orthography directly) to fully phonemic (first mapping each word to a phoneme string). Between these two options lies a continuum of modeling options. We analyze techniques that can improve the accuracy of a graphemic system without requiring significant effort to design or implement. The analysis is performed in the context of the IARPA Babel project, which aims to develop spoken term detection systems for previously unseen languages rapidly, and with minimal human effort. We consider techniques related to letter-to-sound mapping and language-independent syllabification of primarily graphemic systems, and discuss results obtained for six languages: Cebuano, Kazakh, Kurmanji Kurdish, Lithuanian, Telugu and Tok Pisin.

@proceedings{288,
  author = {Marelie Davel and Etienne Barnard and Charl Van Heerden and William Hartman and Damianos Karakos and Richard Schwartz and Stavros Tsakalidis},
  title = {Exploring minimal pronunciation modeling for low resource languages},
  abstract = {Pronunciation lexicons can range from fully graphemic (modeling each word using the orthography directly) to fully phonemic (first mapping each word to a phoneme string). Between these two options lies a continuum of modeling options. We analyze techniques that can improve the accuracy of a graphemic system without requiring significant effort to design or implement. The analysis is performed in the context of the IARPA Babel project, which aims to develop spoken term detection systems for previously unseen languages rapidly, and with minimal human effort. We consider techniques related to letter-to-sound mapping and language-independent syllabification of primarily graphemic systems, and discuss results obtained for six languages: Cebuano, Kazakh, Kurmanji Kurdish, Lithuanian, Telugu and Tok Pisin.},
  year = {2015},
  journal = {Interspeech},
  pages = {538-542},
  address = {Dresden, Germany},
}
Badenhorst J, Davel MH. Synthetic triphones from trajectory-based feature distributions. Pattern Recognition Association of South Africa (PRASA). 2015. doi:10.1109/RoboMech.2015.7359509.

We experiment with a new method to create synthetic models of rare and unseen triphones in order to supplement limited automatic speech recognition (ASR) training data. A trajectory model is used to characterise seen transitions at the spectral level, and these models are then used to create features for unseen or rare triphones. We find that a fairly restricted model (piece-wise linear with three line segments per channel of a diphone transition) is able to represent training data quite accurately. We report on initial results when creating additional triphones for a single-speaker data set, finding small but significant gains, especially when adding additional samples of rare (rather than unseen) triphones.

@proceedings{287,
  author = {Jaco Badenhorst and Marelie Davel},
  title = {Synthetic triphones from trajectory-based feature distributions},
  abstract = {We experiment with a new method to create synthetic models of rare and unseen triphones in order to supplement limited automatic speech recognition (ASR) training data. A trajectory model is used to characterise seen transitions at the spectral level, and these models are then used to create features for unseen or rare triphones. We find that a fairly restricted model (piece-wise linear with three line segments per channel of a diphone transition) is able to represent training data quite accurately. We report on initial results when creating additional triphones for a single-speaker data set, finding small but significant gains, especially when adding additional samples of rare (rather than unseen) triphones.},
  year = {2015},
  journal = {Pattern Recognition Association of South Africa (PRASA)},
  pages = {118-122},
  address = {Port Elizabeth, South Africa},
  isbn = {978-1-4673-7450-7, 978-1-4673-7449-1},
  doi = {10.1109/RoboMech.2015.7359509},
}
  • CSIR
  • DSI
  • Covid-19