Slot-filling and intent detection are well-established tasks in Conversational AI. However, current large-scale benchmarks for these tasks often exclude evaluations of low-resource languages and rely on translations from English benchmarks, thereby predominantly reflecting Western-centric concepts. In this paper, we introduce “INJONGO” - a multicultural, open-source benchmark dataset for 16 African languages with utterances generated by native speakers across diverse domains, including banking, travel, home, and dining. Through extensive experiments, we benchmark fine-tuning multilingual transformer models and prompting large language models (LLMs), and show the advantage of leveraging African-cultural utterances over Western-centric utterances for improving cross-lingual transfer from the English language. Experimental results reveal that current LLMs struggle with the slot-filling task, with GPT-4o achieving an average performance of 26 F1. In contrast, intent detection performance is notably better, with an average accuracy of 70.6%, though it still falls short of fine-tuning baselines. When compared to the English language, GPT-4o and fine-tuning baselines perform similarly on intent detection, achieving an accuracy of approximately 81%. Our findings suggest that LLMs performance is still behind for many low-resource African languages, and more work is needed to further improve their downstream performance.

Despite the widespread adoption of Large language models (LLMs), their remarkable capabilities remain limited to a few high-resource languages. Additionally, many low-resource languages (e.g. African languages) are often evaluated only on basic text classification tasks due to the lack of appropriate or comprehensive benchmarks outside of high-resource languages. In this paper, we introduce IrokoBench—a human-translated benchmark dataset for 17 typologically-diverse low-resource African languages covering three tasks: natural language inference(AfriXNLI), mathematical reasoning(AfriMGSM), and multi-choice knowledge-based QA(AfriMMLU). We use IrokoBench to evaluate zero-shot, few-shot, and translate-test settings(where test sets are translated into English) across 10 open and four proprietary LLMs. Our evaluation reveals a significant performance gap between high-resource languages (such as English and French) and low-resource African languages. We observe a significant performance gap between open and proprietary models, with the highest performing open model, Gemma 2 27B only at 63% of the best-performing proprietary model GPT-4o performance. Machine translating the test set to English before evaluation helped to close the gap for larger models that are English-centric, like Gemma 2 27B and LLaMa 3.1 70B. These findings suggest that more efforts are needed to develop and adapt LLMs for African languages.

The purpose of this article is to demonstrate that the recently developed automated rule-based syllabification system for Sesotho can be used broadly across the officially recognised South African Sotho-Tswana language group encompassing Sepedi, Sesotho and Setswana. We evaluate the automatic syllabification system on 400 words comprising 100 most frequently used words and 100 least-used words in Sepedi and Setswana as evident in the Autshumato corpus publicly available online. It is found that the Sesotho rule-based syllabification system can be used to correctly identify vowel-only syllables, consonant-vowel syllables and consonant-only syllables in Sepedi and Setswana. Among other findings, it has been demonstrated that words with diacritics as in the case of Sepedi are correctly broken down into syllables. We make two main recommendations. First, the rules for syllabification should be updated so that Sepedi diacritics are accommodated. Second, the syllabification system should be updated so that it reflects the broader Sotho-Tswana language group instead of being limited to Sesotho. Further research is needed to ascertain whether the complex consonant [ny] behaves similarly in all three officially recognised Sotho-Tswana languages and evaluate the need for a specific rule for the [ny] digraph.