Generative AI for Named Entity Recognition in Low-Resource Language Nepali

Generative Artificial Intelligence (GenAI), particularly Large Language Models (LLMs), has significantly advanced Natural Language Processing (NLP) tasks, such as Named Entity Recognition (NER), which involves identifying entities like person, location, and organization names in text. LLMs are especially promising for low-resource languages due to their ability to learn from limited data. However, the performance of GenAI models for Nepali, a low-resource language, has not been thoroughly evaluated. This paper investigates the application of state-of-the-art LLMs for Nepali NER, conducting experiments with various prompting techniques to assess their effectiveness. Our results provide insights into the challenges and opportunities of using LLMs for NER in low-resource settings and offer valuable contributions to the advancement of NLP research in languages like Nepali.

Linguistic Taboos and Euphemisms in Nepali

Languages across the world have words, phrases, and behaviors -- the taboos -- that are avoided in public communication considering them as obscene or disturbing to the social, religious, and ethical values of society. However, people deliberately use these linguistic taboos and other language constructs to make hurtful, derogatory, and obscene comments. It is nearly impossible to construct a universal set of offensive or taboo terms because offensiveness is determined entirely by different factors such as socio-physical setting, speaker-listener relationship, and word choices. In this paper, we present a detailed corpus-based study of offensive language in Nepali. We identify and describe more than 18 different categories of linguistic offenses including politics, religion, race, and sex. We discuss 12 common euphemisms such as synonym, metaphor and circumlocution. In addition, we introduce a manually constructed data set of over 1000 offensive and taboo terms popular among contemporary speakers. This in-depth study of offensive language and resource will provide a foundation for several downstream tasks such as offensive language detection and language learning.

Towards A Virtual Assistant That Can Be Taught New Tasks In Any Domain By Its End-Users

The challenge stated in the title can be divided into two main problems. The first problem is to reliably mimic the way that users interact with user interfaces. The second problem is to build an instructible agent, i.e. one that can be taught to execute tasks expressed as previously unseen natural language commands. This paper proposes a solution to the second problem, a system we call Helpa. End-users can teach Helpa arbitrary new tasks whose level of complexity is similar to the tasks available from today's most popular virtual assistants. Teaching Helpa does not involve any programming. Instead, users teach Helpa by providing just one example of a command paired with a demonstration of how to execute that command. Helpa does not rely on any pre-existing domain-specific knowledge. It is therefore completely domain-independent. Our usability study showed that end-users can teach Helpa many new tasks in less than a minute each, often much less.