New text analytics plugin painlessly delivers rich, faceted search
An API key and a line of code is all it takes to speed your research, enhance voice of the customer systems, automate content recommendations and more.
Rosette API for Elasticsearch
We launched Rosette API to put text analytics in more hands. Through the Rosette Document Enrichment plugin for Elasticsearch, we extend Rosette’s reach through our partners like Elastic who have millions of users. With Elastic’s powerful set of solutions, from ingestion to parsing and visualization, Rosette’s Document Enrichment plugin delivers better search results, going beyond simple keyword search to discover new insights into your data.
From raw text to insights
Enrich your Elasticsearch index records with fields for entities, sentiment, name translation, language identification (55 languages), and categorization facets with Rosette API’s Document Enrichment plugin time. With extensive language support and customization parameters available on-premises, Rosette allows you to find the information you need from document collections of any size.
Elasticsearch developers choose which facets to add straight from the cloud. Simply configure the processors you need and Rosette API delivers: offsets, language identification, entity extraction, categorization, name translation, and sentiment analysis. For example, a search of our demo database for “Hamilton” returns 18 articles, concisely summarized below:
Abstracts of each article include high-level data such as the language, category, sentiment, and recognized entities and entity types:
Working with multilingual documents? The plugin also includes automated translation of extracted entities, like Marine Le Pen shown below:
While each of these text analytics tools are available individually through Rosette API, we’ve made the work easy for you by combining them in a comprehensive and easy-to-use package that’s unique to the market.
Filter results by sentiment, entity, category
The Document Enrichment plugin allows users to dramatically enhance their search capabilities by adding facets for offsets, entities, sentiment, name translation, language, and category. You can easily make multifaceted queries for deep analysis, drilling down to just the documents that are most interesting to you.
For example, limiting our results from the “Hamilton” search above to include only documents with a positive sentiment decreases our results from 18 to 7:
You can also search for specific entities – not just their name as a keyword – which will return documents that may have misspelled the entity, but are close enough to still be a match. For example, the typo “Donald Thump” was still recognized by an entity search:
And “Wladimir Putin” is still recognized in a search for “Vladimir Putin:”
Document enrichment in the wild
Document enrichment enhances the search experience. Once the documents in your Elasticsearch index have been enriched new queries can be made to answer complex questions.
- Looking to automate a content recommendation engine? Use Rosette Document Enrichment to quickly locate any documents in your index related to sports and provide your users the content they want.
- Performing market research on the success of a new product? Multifaceted queries can allows you to access the “voice of customer” on an entirely new level. For example, filter your customer reviews to find negative feedback that specifically mention the product in question, enabling product improvements and more satisfied customers.
- Researching how sentiment towards a public figure varies in different countries around the world for an article? Quickly find the sentiment break down of articles mentioning Angela Merkel, the Pope, or Oprah in English, Japanese or Spanish.
Document enrichment can improve the speed and efficacy of anyone who works with a large database of text-based documents. While many of the features included within the Document Enrichment plugin are available through multiple text analytics providers and open source offerings, the Rosette plugin simplifies the process by automating the analysis of your Elasticsearch index with minimal coding.
