Skill Classifiers

Fast-Text Classifier: Best for text classification tasks requiring high speed and scalability with moderate accuracy. It is efficient for large datasets and performs well in classifying short texts like user comments or tweets.

BiLSTM Classifier: Ideal for sequence-based data, such as sentiment analysis, named entity recognition, or time-series data, where understanding the context before and after a word is crucial. It excels in handling long-range dependencies in text.

BM25 Classifier: Used primarily in information retrieval tasks, such as search engines or document ranking, where relevance scoring based on term frequency and inverse document frequency (TF-IDF) is essential for matching queries with documents.

NLI Classifier: Best for natural language inference tasks, where determining the relationship between two sentences (e.g., entailment, contradiction, or neutral) is required. Useful in contexts like question answering and dialogue systems.

LLM Classifier: Suitable for tasks requiring deep understanding and generalization across diverse language contexts, such as text generation, summarization, and complex classification tasks. LLMs (Large Language Models) are highly accurate for nuanced language tasks but resource-intensive.