Search Configuration - STILINGUE English May 20, 2024 13:31 Updated Index: How it works Artificial Intelligence Engine - STILINGUE English How it works Learn how to configure your search to collect relevant publications in English. STILINGUE English offers everything you already know from our platform, now available for analysis and collection of publications in English. When creating a new English search panel in STILINGUE, it can be configured according to the researched context, in addition to entering all necessary terms for collection in Groups, classifying in Themes, Tags, and other necessary information, just like in searches in Portuguese and Spanish. For more information on this process, read our basic Search Configuration guide. Select the desired language for your Context and also for your Groups and Themes: Thinking specifically about English, it is important to enter terms specific to this language, respecting elements such as spelling, variations, and other differences with other languages. If you choose to include terms in Portuguese and/or Spanish, include the correct terms from the respective languages when necessary. Publications in English will be collected only via Listening, and will be available for analysis in all graphs in this area of the platform. Artificial Intelligence Engine - STILINGUE English STILINGUE English offers everything you already know from our platform, now available for analysis and collection of publications in English. Click here and learn more about the product. To collect data in the English language and classify it accurately, STILINGUE English relies on a specific artificial intelligence engine designed for this language. This engine processes the collected information separately from other data in Portuguese or Spanish, ensuring that each language and its peculiarities are analyzed and understood. What Does an Artificial Intelligence Engine Do? To optimize access to millions of pieces of information and conversations in Portuguese, STILINGUE uses Artificial Intelligence (AI) to provide real-time accurate data, through exclusive STILINGUE natural language processing (SNLP). Refer to the guide for detailed information on how natural language processing works. SNLP is the exclusive version created by the STILINGUE team of experts to translate written language into a computational version through language processing. This functionality can be found in various areas of the platform: in automatic sentiment classification, term relevance ranking, automatic gender detection, automatic language filtering, and AAA Ranking. The experience that STILINGUE has gained in creating the best AI engines for collecting and analyzing publications in Portuguese and Spanish has been applied to the development of a new AI, exclusively focused on the English language. This new engine already boasts an estimated accuracy rate of 70%, when compared to other international benchmarks. Note: the mentioned results may always depend on the sample and data used for comparison. STILINGUE English in practice Language selection is made during the Search Configuration, with options including English, Spanish, Portuguese, or a combination of languages. Refer to the guide for more information. In the example below, we will demonstrate the differences in sentiment analysis ratings for an English post by different AI engines: "STILINGUE artificial intelligence engine specially developed for EN prequalifies and processes data, freeing more time for analysis composition and insights generation." English: Positive Spanish: Negative Portuguese: Negative We can observe that the engine specialized in English provided a much more accurate classification of the post when compared to those dedicated to Spanish and Portuguese. This accuracy implies a reduction in classification errors and also in the quantity of unclassified and neutral posts present in the research. When selecting languages in the Search Configuration, the language of each post is identified and directed to the specific AI engine for that language, ensuring that each language and its peculiarities are understood. Note: Language classification in textless posts - where the content consists only of links, emojis, user mentions, numbers, email, or insufficient letters - will be conducted through language information available in links, mentions, or hashtags. For more information, visit the discussion on the subject at our community or videos on our channel. 😃 Related articles FAQs - Listening Themes Sending Active Messages to WhatsApp via Growth Reels - Instagram STILINGUE Spanish