The term low-resourced has been tossed around in the field of natural language processing to a degree that almost any language that is not English can be called "low-resourced"; sometimes even just for the sake of making a mundane or mediocre paper appear more interesting and insightful. In a field where English is a synonym for language and low-resourced is a synonym for anything not English, calling endangered languages low-resourced is a bit of an overstatement. In this paper, I inspect the relation of the endangered with the low-resourced from my own experiences.

An increase in the use of smartphones has laid to the use of the internet and social media platforms. The most commonly used social media platforms are Twitter, Facebook, WhatsApp and Instagram. People are sharing their personal experiences, reviews, feedbacks on the web. The information which is available on the web is unstructured and enormous. Hence, there is a huge scope of research on understanding the sentiment of the data available on the web. Sentiment Analysis (SA) can be carried out on the reviews, feedbacks, discussions available on the web. There has been extensive research carried out on SA in the English language, but data on the web also contains different other languages which should be analyzed. This paper aims to analyze, review and discuss the approaches, algorithms, challenges faced by the researchers while carrying out the SA on Indigenous languages.

Using a safe subset of C++ is a promising direction for increasing the safety of the programming language while maintaining its performance and productivity. In this paper, we examine how close existing C/C++ code is to conforming to a safe subset of C++. We examine the rules presented in existing safe C++ standards and safe C++ subsets. We analyze the code characteristics of 5.8 million code samples from the Exebench benchmark suite and 5 modern C++ applications using a static analysis tool. We find that raw pointers, unsafe casts, and unsafe library functions are used in both C++ code at large and modern C++ applications. In general, C++ code at large does not differ much from modern C++ code, and continued work will be required to transition from existing C/C++ code to a safe subset of C++.

The web ecosystem is rapidly evolving with changing business and functional models. Cloud platforms are available in a SaaS, PaaS and IaaS model designed around commoditized Linux based servers. 10 billion users will be online and accessing the web and its various content. The industry has seen a convergence around IP based technology. Additionally, Linux based designs allow for a system wide profiling of application characteristics. The customer is an OEM who provides Linux based servers for telecom solutions. The end customer will develop business applications on the server. Customers are interested in a latency profiling mechanism which helps them to understand how the application behaves at run time. The latency profiler is supposed to find the code path which makes an application block on I/O, and other synchronization primitives. This will allow the customer to understand the performance bottleneck and tune the system and application parameters.

In mathematics, the Riemann hypothesis is a conjecture that the Riemann zeta function has its zeros only at the negative even integers and complex numbers with real part 1/2. Many consider it to be the most important unsolved problem in pure mathematics. It is one of the seven Millennium Prize Problems selected by the Clay Mathematics Institute to carry a US 1,000,000 prize for the first correct solution. We prove the Riemann hypothesis using the Complexity Theory. Number theory is a branch of pure mathematics devoted primarily to the study of the integers and integer-valued functions. The Goldbach's conjecture is one of the most important and unsolved problems in number theory. Nowadays, it is one of the open problems of Hilbert and Landau. We show the Goldbach's conjecture is true using the Complexity Theory as well. An important complexity class is 1NSPACE(S(n)) for some S(n). These mathematical proofs are based on if some unary language belongs to 1NSPACE(S(log n)), then the binary version of that language belongs to 1NSPACE(S(n)) and vice versa.

We present improvements in n-best rescoring of code-switched speech achieved by n-gram augmentation as well as optimised pretraining of long short-term memory (LSTM) language models with larger corpora of out-of-domain monolingual text. In addition, we consider the application of large pretrained transformer-based architectures. Our experimental evaluation is performed on an under-resourced corpus of code-switched speech comprising four bilingual code-switched sub-corpora, each containing a Bantu language (isiZulu, isiXhosa, Sesotho, or Setswana) and English. We find in our experiments that, by combining n-gram augmentation with the optimised pretraining strategy, speech recognition errors are reduced for each individual bilingual pair by 3.51% absolute on average over the four corpora. Importantly, we find that even speech recognition at language boundaries improves by 1.14% even though the additional data is monolingual. Utilising the augmented n-grams for lattice generation, we then contrast these improvements with those achieved after fine-tuning pretrained transformer-based models such as distilled GPT-2 and M-BERT. We find that, even though these language models have not been trained on any of our target languages, they can improve speech recognition performance even in zero-shot settings. After fine-tuning on in-domain data, these large architectures offer further improvements, achieving a 4.45% absolute decrease in overall speech recognition errors and a 3.52% improvement over language boundaries. Finally, a combination of the optimised LSTM and fine-tuned BERT models achieves a further gain of 0.47% absolute on average for three of the four language pairs compared to M-BERT. We conclude that the careful optimisation of the pretraining strategy used for neural network language models can offer worthwhile improvements in speech recognition accuracy even at language switches, and that much larger state-of-the-art architectures such as GPT-2 and M-BERT promise even further gains.

So-called ‘zero’ or ‘null’ tenses have often been characterized as functionally deficient forms, deprived of any inherent content. In this paper, we will focus on the semantics of a morpho-phonologically null inflectional verbal paradigm in Anindilyakwa (Groote Eylandt, N.T., Australia, which is both temporally and aspectually underspecified. Through a quantitative corpus study conducted in the paper, we establish that ‘zero inflection’ in this language, contra prior works on such tenses in general (e.g. Bybee 1990) and in Anindilyakwa in particular (Bednall 2019), presents various degrees of sensitivity to traditional Vendlerian aspectual parameters. We show that while telicity is not a significant predictor for the temporal interpretation of zero-inflected Anindilyakwa verbs, and dynamicity is a good but not very good predictor, only a very broad opposition between change-of-state (including qua boundedness) and non-change-of-state, or perfective/imperfective, gives very significant biases towards past vs. present anchoring. We also show that atomic telicity is the only categorical Aktionsart predictor for temporal anchoring in this context correctly predict the temporal anchoring of such verbs, and stativity is not biased towards present interpretations, thereby questioning currently received typological theories of the semantics of so-called ‘zero-tenses’ / aspectuo-temporally underspecified tenses.

This paper investigates the meaning of a specific intonation contour found in the Northern Australian language Iwaidja called Linear Lengthening Intonation (LLI). Using an experimental field work approach, we analysed approximately 4,000 utterances. We demonstrate that the semantics of LLI is broadly event-quantificational as well as temporally scalar. LLI imposes aspectual selectional restrictions on the verbs it combines with (they must be durative, i.e. cannot describe ‘punctual’, atomic events), and requires the event description effected by said verbs to exceed a contextually-determined relative scalar meaning (e.g., a ‘typical duration’ à la (Tatevosov 2008)). Iwaidja differs from other Northern Australian languages with similar intonation patterns (see e.g. (Bishop 2002: 2002; Simard 2013)), in that it does not seem to have any argument NP-related incremental or event scalar meaning. This suggests that LLI is a decidedly grammatical, language-specific device; not a purely iconic kind of expression (even though it also possibly has an iconic dimension).

It is time that universities reexamine what is meant by globalization. Contemporary researchers in science and the humanities (Critchley, Chomsky, Mumford, Ostrom, Eisenstein, Ferry, Orr, Shiva, Klein, Margulis, Meadows, Capra and Tolba, just to name a few) have aptly redefined the concept of « world » as a biological and cultural ecosystem. This paper seeks ways to integrate the theory and practice of eco-citizenship into various cross-disciplinary aspects of higher education, with a focus on curricular adjustments that may be steered by World Languages and Cultures programs. While "global citizenship" is still often understood today as a form of supranational citizenship that may find its actualization through the valuable, yet often arrested efforts of the United Nations, or as the individualistic result of a neoliberal economic emancipation of markets and capital throughout the world, this notion must rather be embedded within a radically cultural, natural and ethical bedrock from which a more potent world citizenry will stem. Departments of World Languages and Cultures and cultures are ideally positioned in the academic landscape to foster the development of a greater eco-civic and biospheric awareness that can permeate new curricular orientations of universities in the US and abroad.

With the existence of several programming languages such as C/C++, Java, C#, LISP, Prolog, Python, Simula, F#, Go, Haskell, Scala, Ruby, Dart, Swift, Groovy etc. and diverse paradigms like structured, object-oriented, list, aspect-oriented, service-oriented, web, mobile and logic programming, there is a need to perform an exhaustive comparative analysis of diverse compilers and environments before making a choice of implementation technology in software engineering. Optimization of compilers helps to reduce execution time by making use of high speed processor registers, thereby, eliminating redundant computation. This paper reports some series of performance analysis done with some popular programming languages including Java, C++, Python and PHP. Programs involving recursive and iterative functions like factorial of large numbers and binary search of large arrays were run on the various platforms with the execution time recorded in milliseconds and represented in a chart. This can aid in making a selection of the appropriate language to use for a given application domain.

The mathematical structure underlying a sentence is made of two independent processing units in series. The conditional probability density function of sentence length (measured in words) recordable in the extended short–term memory (E–STM) buffer made of CF cells is Gaussian. The number of sentences that authors of Italian and English Literatures have written is compared to that theoretically available to them. The multiplicity factor α describes this comparison. α>1 is more likely than α<1 and often α≫1, therefore the same sentence pattern is reused many times. These texts are easier to read. Very few novels show α<1. In these cases, there are enough diverse sentence patters to convey meaning, but most of them are not used. Texts are more difficult to read, as measured by a universal readability index GU. A mismatch index IM synthetically describes the process. α, GU. and IM increase with year of publication. Future work will concern other literatures – including ancient ones as Greek and Latin – to confirm what, in our opinion, is general feature of human mind.

Through the use of Smart Sensor Networks (SSN), Marine Observatories (MO) provide continuous ocean monitoring. Deployed sensors may not perform as intended due to the heterogeneity of SSN devices’ hardware and software when combined with the Internet. Hence, SSN are regarded as complex distributed systems. As such, SSN designers will encounter challenges throughout the design phase related to time, complexity, sharing diverse domain experiences (viewpoints), and ensuring optimal performance for the deployed SSN. To this end, we describe in this article how we extended our previously proposed ArchiMO meta-model and design tool by extending an Enterprise Architecture Framework based on the ArchiMate Modeling Language. This extension proposes the incorporation of new Underwater Environmental Constraints (UEC) for SSN in ArchiMO which is coupled with the fusion of sensors with Artificial Intelligence. This serves as the basis for generating a new version of our ArchiMO design tool, which incorporates the new added constraints and invokes the AI created Database. To illustrate our proposal, we use the newly generated ArchiMO to create a model in the MO domain. Furthermore, we use our self-developed domain-specific model compiler to produce the relevant simulation code. Throughout the design phase, our approach contributes to handle and control the uncertainties and variances of the provided quality of service that may occur during the performance of the SSN, hence reducing the design activity’s complexity and time. It provides a way to share the different viewpoints of the designers in the domain of SSN.

We study the short−term memory capacity of ancient readers of the original New Testament written in Greek, of its translations to Latin and modern languages. To model the short–term capacity, we have considered the number of words per interpunctions, the “word interval” , because this parameter can model how the human mind memorizes “chunks” of information. Since can be calculated for any alphabetical text, we can perform experiments − otherwise impossible − with ancient readers by studying the literary works they used to read. The “experiments” compare the of texts of a language/translation to those of another language/translation by measuring the minimum average probability of finding joint readers (those who can read both texts because of their similar short–-term memory capacity) and by defining an “overlap index”. We also define a population of universal readers who can read any the New Testament written in alphabetical language. More than 50% of the readers of specific languages overlap with the universal readers with probability . Future work is vast, with many research tracks, because alphabetical literatures are very large and allow many experiments, such as comparing authors, translations or even texts written by artificial intelligence tools.

Recent improvements in deepfake creation have made deepfake videos more realistic. Moreover, open-source software has made deepfake creation more accessible, which reduces the barrier to entry for deepfake creation. This could pose a threat to the people’s privacy. There is a potential danger if the deepfake creation techniques are used by people with an ulterior motive to produce deepfake videos of world leaders to disrupt the order of countries and the world. Therefore, research into the automatic detection of deepfaked media is essential for public security. In this work, we propose a deepfake detection method using upper body language analysis. Specifically, a many-to-one LSTM network was designed and trained as a classification model for deepfake detection. Different models were trained by varying the hyperparameters to build a final model with benchmark accuracy. We achieved 94.39% accuracy on the deepfake test set. The experimental results showed that upper body language can effectively detect deepfakes.