Lexalytics is a text analytics and natural language processing platform provided in the form of APIs and software libraries, mainly used to transform complex text documents into data, insights, and business value, and can be integrated into products, platforms, or applications.
Main capabilities include: sentiment analysis, determining whether text is positive, negative, or neutral; entity extraction, identifying people, places, dates, companies, products, job titles, and more; intent detection, determining the intent expressed by customers or reviewers; document classification, grouping documents into customizable categories; it also provides prebuilt industry configurations to improve sentiment accuracy, topic detection, classification, and other results.
In terms of usage, the official website presents three forms: low-level text analytics and NLP software libraries for data scientists and architects, suitable for scenarios that require full access to the underlying technology or on-premises deployment for security and privacy reasons; a RESTful API that encapsulates Salience capabilities, equipped with graphical configuration and user management tools, which can be integrated into cloud-based enterprise data analysis infrastructure or provided to one's own customers; and a complete solution based on the Semantria API, used to store, manage, and analyze unstructured text documents, and to visualize results and share findings through interactive dashboards. Deployment supports any combination of on-premises, private cloud, hybrid cloud, and public cloud.
Applicable scenarios include consumer feedback and intent monitoring in industries such as hotels, dining, and retail; in the healthcare field, drugs, symptoms, diseases, dosages, anatomical terms, specialties, and 250 ICD-10 codes mapped to standardized names; as well as voice of the employee and others. The official website states that it has more than 19 years of development history, continuously adds new libraries, configurations, and models, has deployed hundreds of machine learning models, and can also train custom micromodels on demand. In terms of configuration, it supports defining custom entities, creating query topics, building taxonomies, adding blacklists, and more.
In terms of limitations, publicly available information does not provide specific pricing, free trials, quotas, SLAs, or complete technical documentation details; the depth of integration with third-party systems, accuracy metrics, and the specific performance of supported languages are not yet provided.
