There's a constant need to reason about diverse cultures all over the world. Past cultural-reasoning research has focused primarily on techniques to organize, catalog, and reason about cultural and historical artifacts of the kind typically stored in a museum. This is extremely valuable. However, the term "cultural reasoning" as we use it in the previous examples (and in this article) focuses on understanding how different cultural groups today make decisions and what factors those decisions are based on. An architecture that supports cultural reasoning should, for example, be able to pinpoint characteristics that differentiate organizations engaging political action within legitimate frameworks from those engaging in violence and terror. Key in all this is that cultural reasoning must go hand in hand with environmental reasoning. We believe that any architecture to support cultural reasoning about a given group, political entity, business, or religious organization should contain these components: 1) a semantic Web extraction engine to elicit data about the organization, 2) an opinion-mining engine that captures the organization's opinions, 3) an algorithm to correlate environmental variables with actions that the organization takes, and 4) a simulation or game environment within which analysts and users can see what the organization has done and what it might do in hypothetical situations

In recent years, the extraction of opinions and information from user-generated text has attracted a lot of interest, largely due to the unprecedented volume of content in Social Media. However, social researchers face some issues in adopting cutting-edge tools for these tasks, as they are usually behind commercial APIs, unavailable for other languages than English, or very complex to use for non-experts. To address these issues, we present pysentimiento, a comprehensive multilingual Python toolkit designed for opinion mining and other Social NLP tasks. This open-source library brings state-of-the-art models for Spanish, English, Italian, and Portuguese in an easy-to-use Python library, allowing researchers to leverage these techniques. We present a comprehensive assessment of performance for several pre-trained language models across a variety of tasks, languages, and datasets, including an evaluation of fairness in the results.

In recent years, the extraction of opinions and information from user-generated text has attracted a lot of interest, largely due to the unprecedented volume of content in Social Media. However, social researchers face some issues in adopting cutting-edge tools for these tasks, as they are usually behind commercial APIs, unavailable for other languages than English, or very complex to use for non-experts. To address these issues, we present pysentimiento, a comprehensive multilingual Python toolkit designed for opinion mining and other Social NLP tasks. This open-source library brings state-of-the-art models for Spanish, English, Italian, and Portuguese in an easy-to-use Python library, allowing researchers to leverage these techniques. We present a comprehensive assessment of performance for several pre-trained language models across a variety of tasks, languages, and datasets, including an evaluation of fairness in the results.