scholarly journals Developing a Shared Task for Speech Processing on Endangered Languages

2021 ◽  
Vol 1 (2) ◽  
Author(s):  
Gina-Anne Levow ◽  
Emily P. Ahn ◽  
Emily M. Bender
Keyword(s):  
Language Processing ◽  
Speech Processing ◽  
Data Sets ◽  
Language Documentation ◽  
Data Set ◽  
Endangered Language ◽  
Speech Communities ◽  
Resource Requirements ◽  
The Creation ◽  
Speech And Language Processing

Advances in speech and language processing have enabled the creation of applications that could, in principle, accelerate the process of language documentation, as speech communities and linguists work on urgent language documentation and reclamation projects. However, such systems have yet to make a significant impact on language documentation, as resource requirements limit the broad applicability of these new techniques. We aim to exploit the framework of shared tasks to focus the technology research community on tasks which address key pain points in language documentation. Here we present initial steps in the implementation of these new shared tasks, through the creation of data sets drawn from endangered language repositories and baseline systems to perform segmentation and speaker labeling of these audio recordings—important enabling steps in the documentation process. This paper motivates these tasks with a use case, describes data set curation and baseline systems, and presents results on this data. We then highlight the challenges and ethical considerations in developing these speech processing tools and tasks to support endangered language documentation.

Word recognition in aphasia

2007 ◽  
pp. 140-156
Author(s):  
Sheila Blumstein
Keyword(s):  
Word Recognition ◽  
Language Processing ◽  
Speech Processing ◽  
Current Knowledge ◽  
Recognition System ◽  
Speech And Language ◽  
Neural Basis ◽  
Sentential Context ◽  
The Face ◽  
Speech And Language Processing

This article reviews current knowledge about the nature of auditory word recognition deficits in aphasia. It assumes that the language functioning of adults with aphasia was normal prior to sustaining brain injury, and that their word recognition system was intact. As a consequence, the study of aphasia provides insight into how damage to particular areas of the brain affects speech and language processing, and thus provides a crucial step in mapping out the neural systems underlying speech and language processing. To this end, much of the discussion focuses on word recognition deficits in Broca's and Wernicke's aphasics, two clinical syndromes that have provided the basis for much of the study of the neural basis of language. Clinically, Broca's aphasics have a profound expressive impairment in the face of relatively good auditory language comprehension. This article also considers deficits in processing the sound structure of language, graded activation of the lexicon, lexical competition, influence of word recognition on speech processing, and influence of sentential context on word recognition.

scholarly journals Synchronous, but not Entrained: Exogenous and Endogenous Cortical Rhythms of Speech and Language Processing

2019 ◽  
Author(s):  
Lars Meyer ◽  
Yue Sun ◽  
Andrea E. Martin
Keyword(s):  
Language Development ◽  
Language Processing ◽  
Speech Processing ◽  
Oscillatory Activity ◽  
Syntactic Structures ◽  
Acoustic Information ◽  
Cortical Rhythms ◽  
Abstract Representations ◽  
New Research ◽  
Speech And Language Processing

Research into speech processing is often focused on a phenomenon termed ‘entrainment’, whereby the cortex shadows rhythmic acoustic information with oscillatory activity. Entrainment has been observed to a range of rhythms present in speech; in addition, synchronicity with abstract information (e.g., syntactic structures) has been observed. Entrainment accounts face two challenges: First, speech is not exactly rhythmic; second, synchronicity with representations that lack a clear acoustic counterpart has been described. We propose that apparent entrainment does not always result from acoustic information. Rather, internal rhythms may have functionalities in the generation of abstract representations and predictions. While acoustics may often provide punctate opportunities for entrainment, internal rhythms may also live a life of their own to infer and predict information, leading to intrinsic synchronicity—not to be counted as entrainment. This possibility may open up new research avenues in the psycho– and neurolinguistic study of language processing and language development.

Intelligent Speech Processing Technique for Suspicious Voice Call Identification Using Adaptive Machine Learning Approach

2021 ◽  
pp. 372-380
Author(s):  
Dhilip Kumar ◽  
Swathi P. ◽  
Ayesha Jahangir ◽  
Nitesh Kumar Sah ◽  
Vinothkumar V.
Keyword(s):  
Mobile Phones ◽  
Language Processing ◽  
Speech Processing ◽  
Personal Information ◽  
Processing Technique ◽  
Learning Approach ◽  
Data Set ◽  
Intelligent Algorithms ◽  
Machine Learning Approach ◽  
Incoming Call

With recent advances in the field of data, there are many advantages of speedy growth of internet and mobile phones in the society, and people are taking full advantage of them. On the other hand, there are a lot of fraudulent happenings everyday by stealing the personal information/credentials through spam calls. Unknowingly, we provide such confidential information to the untrusted callers. Existing applications for detecting such calls give alert as spam to all the unsaved numbers. But all calls might not be spam. To detect and identify such spam calls and telecommunication frauds, the authors developed the application for suspicious call identification using intelligent speech processing. When an incoming call is answered, the application will dynamically analyze the contents of the call in order to identify frauds. This system alerts such suspicious calls to the user by detecting the keywords from the speech by comparing the words from the pre-defined data set provided to the software by using intelligent algorithms and natural language processing.

scholarly journals Unassisted Noise-Reduction of Chemical Reactions Data Sets

2021 ◽  
Author(s):  
Alessandra Toniato ◽  
Philippe Schwaller ◽  
Antonio Cardinale ◽  
Joppe Geluykens ◽  
Teodoro Laino
Keyword(s):  
Noise Reduction ◽  
Chemical Reactions ◽  
Language Processing ◽  
Prediction Models ◽  
Open Data ◽  
Data Sets ◽  
Data Set ◽  
Intelligence Models ◽  
Artificial Intelligence Models ◽  
Jensen Shannon Divergence

<p>Existing deep learning models applied to reaction prediction in organic chemistry can reach high levels of accuracy (> 90% for Natural Language Processing-based ones).</p><p>With no chemical knowledge embedded than the information learnt from reaction data, the quality of the data sets plays a crucial role in the performance of the prediction models. While human curation is prohibitively expensive, the need for unaided approaches to remove chemically incorrect entries from existing data sets is essential to improve artificial intelligence models' performance in synthetic chemistry tasks. Here we propose a machine learning-based, unassisted approach to remove chemically wrong entries from chemical reaction collections. We applied this method to the collection of chemical reactions Pistachio and to an open data set, both extracted from USPTO (United States Patent Office) patents. Our results show an improved prediction quality for models trained on the cleaned and balanced data sets. For the retrosynthetic models, the round-trip accuracy metric grows by 13 percentage points and the value of</p><p>the cumulative Jensen Shannon divergence decreases by 30% compared to its original record. The coverage remains high with 97%, and the value of the class-diversity is not affected by the cleaning. The proposed strategy is the first unassisted rule-free technique to address automatic noise reduction in chemical data sets.</p>

scholarly journals Neurocomputational Models of Language Processing

2021 ◽  
Vol 8 (1) ◽  
Author(s):  
John T. Hale ◽  
Luca Campanelli ◽  
Jixing Li ◽  
Shohini Bhattasali ◽  
Christophe Pallier ◽  
...  
Keyword(s):  
Language Processing ◽  
Computational Models ◽  
Annual Review ◽  
Publication Date ◽  
Data Sets ◽  
Linguistic Features ◽  
Neural Signals ◽  
Data Set ◽  
Interpretable Models ◽  
Compare And Contrast

Efforts to understand the brain bases of language face the Mapping Problem: At what level do linguistic computations and representations connect to human neurobiology? We review one approach to this problem that relies on rigorously defined computational models to specify the links between linguistic features and neural signals. Such tools can be used to estimate linguistic predictions, model linguistic features, and specify a sequence of processing steps that may be quantitatively fit to neural signals collected while participants use language. Progress has been helped by advances in machine learning, attention to linguistically interpretable models, and openly shared data sets that allow researchers to compare and contrast a variety of models. We describe one such data set in detail in the Supplementary Appendix. Expected final online publication date for the Annual Review of Linguistics, Volume 8 is January 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

scholarly journals Microsoft Academic Graph: When experts are not enough

2020 ◽  
Vol 1 (1) ◽  
pp. 396-413 ◽  
Author(s):  
Kuansan Wang ◽  
Zhihong Shen ◽  
Chiyuan Huang ◽  
Chieh-Han Wu ◽  
Yuxiao Dong ◽  
...  
Keyword(s):  
Natural Language ◽  
Language Processing ◽  
Open Data ◽  
Data Sets ◽  
Data Set ◽  
Scholarly Publications ◽  
Scholarly Communications ◽  
Microsoft Academic ◽  
Commercial Applications ◽  
Ongoing Project

An ongoing project explores the extent to which artificial intelligence (AI), specifically in the areas of natural language processing and semantic reasoning, can be exploited to facilitate the studies of science by deploying software agents equipped with natural language understanding capabilities to read scholarly publications on the web. The knowledge extracted by these AI agents is organized into a heterogeneous graph, called Microsoft Academic Graph (MAG), where the nodes and the edges represent the entities engaging in scholarly communications and the relationships among them, respectively. The frequently updated data set and a few software tools central to the underlying AI components are distributed under an open data license for research and commercial applications. This paper describes the design, schema, and technical and business motivations behind MAG and elaborates how MAG can be used in analytics, search, and recommendation scenarios. How AI plays an important role in avoiding various biases and human induced errors in other data sets and how the technologies can be further improved in the future are also discussed.

scholarly journals Improving Basic Natural Language Processing Tools for the Ainu Language

Information ◽  
2019 ◽  
Vol 10 (11) ◽  
pp. 329
Author(s):  
Karol Nowakowski ◽  
Michal Ptaszynski ◽  
Fumito Masui ◽  
Yoshio Momouchi
Keyword(s):  
Language Processing ◽  
Positive Impact ◽  
Word Segmentation ◽  
Data Set ◽  
Part Of Speech Tagging ◽  
Endangered Language ◽  
Part Of Speech ◽  
Overall Performance ◽  
Speech Tagging ◽  
Northern Japan

Ainu is a critically endangered language spoken by the native inhabitants of northern Japan. This paper describes our research aimed at the development of technology for automatic processing of text in Ainu. In particular, we improved the existing tools for normalizing old transcriptions, word segmentation, and part-of-speech tagging. In the experiments we applied two Ainu language dictionaries from different domains (literary and colloquial) and created a new data set by combining them. The experiments revealed that expanding the lexicon had a positive impact on the overall performance of our tools, especially with test data unrelated to any of the training sets used.

scholarly journals Data augmentation strategies to improve reaction yield predictions and estimate uncertainty

2020 ◽  
Author(s):  
Philippe Schwaller ◽  
Alain C. Vaucher ◽  
Teodoro Laino ◽  
Jean-Louis Reymond
Keyword(s):  
Chemical Reactions ◽  
Language Processing ◽  
Data Augmentation ◽  
Test Time ◽  
Reaction Yield ◽  
Small Data ◽  
Data Sets ◽  
Prediction Errors ◽  
Data Set ◽  
Uncertainty Estimates

Chemical reactions describe how precursor molecules react together and transform into products. The reaction yield describes the percentage of the precursors successfully transformed into products relative to the theoretical maximum. The prediction of reaction yields can help chemists navigate reaction space and accelerate the design of more effective routes. Here, we investigate the best-studied high-throughput experiment data set and show how data augmentation on chemical reactions can improve yield predictions' accuracy, even when only small data sets are available. Previous work used molecular fingerprints, physics-based or categorical descriptors of the precursors. In this manuscript, we fine-tune natural language processing-inspired reaction transformer models on different augmented data sets to predict yields solely using a text-based representation of chemical reactions. When the random training sets contain 2.5% or more of the data, our models outperform previous models, including those using physics-based descriptors as inputs. Moreover, we demonstrate the use of test-time augmentation to generate uncertainty estimates, which correlate with the prediction errors.

scholarly journals CHANGES AFFECTING GENERALIZATION OF LAND COVER FEATURES IN A SMALLER SCALE

2012 ◽  
Vol 38 (3) ◽  
pp. 98-105 ◽  
Author(s):  
Lina Papšienė ◽  
Kęstutis Papšys
Keyword(s):  
Land Cover ◽  
Spatial Data ◽  
Data Sets ◽  
Data Set ◽  
The Creation ◽  
Spatial Data Sets

Reference spatial data sets represent the least changing natural and anthropogenic features of terrine. As a rule, such data are stored in different scales and most frequently updated consequently starting with a spatial data set of a larger scale (usually base scale) thus later performing an update of data in smaller scales. The generalization of features in a larger scale is one of the major processes employed in the creation and update of spatial data of a smaller scale. In order to effectively carry out works, it is recommended to use automatic procedures and generalization only in those cases when changes in features are significant, i.e. affect the update of features in a smaller scale. The article discusses the relation between changes in polygon features (identify land cover territories in a base spatial data set) and different generalization processes as well as the evaluation of significance of likely changes.

scholarly journals Data augmentation strategies to improve reaction yield predictions and estimate uncertainty

2020 ◽  
Author(s):  
Philippe Schwaller ◽  
Alain C. Vaucher ◽  
Teodoro Laino ◽  
Jean-Louis Reymond
Keyword(s):  
Chemical Reactions ◽  
Language Processing ◽  
Data Augmentation ◽  
Test Time ◽  
Reaction Yield ◽  
Small Data ◽  
Data Sets ◽  
Prediction Errors ◽  
Data Set ◽  
Uncertainty Estimates

Chemical reactions describe how precursor molecules react together and transform into products. The reaction yield describes the percentage of the precursors successfully transformed into products relative to the theoretical maximum. The prediction of reaction yields can help chemists navigate reaction space and accelerate the design of more effective routes. Here, we investigate the best-studied high-throughput experiment data set and show how data augmentation on chemical reactions can improve yield predictions' accuracy, even when only small data sets are available. Previous work used molecular fingerprints, physics-based or categorical descriptors of the precursors. In this manuscript, we fine-tune natural language processing-inspired reaction transformer models on different augmented data sets to predict yields solely using a text-based representation of chemical reactions. When the random training sets contain 2.5% or more of the data, our models outperform previous models, including those using physics-based descriptors as inputs. Moreover, we demonstrate the use of test-time augmentation to generate uncertainty estimates, which correlate with the prediction errors.

Sign in / Sign up

Export Citation Format

Share Document