SAS® Text Analytics updates: turning big data volume into value
SAS® Analytics unearths insight, value through active learning
Analytics 2012, LAS VEGAS (08. Okt. 2012) – Companies inundated with data are desperate to turn volume into value. Many use predictive modeling to glean insights from structured data in databases. Unstructured data is plentiful and can also hold valuable insights. However, manual categorization of unstructured text can be costly and time-consuming. SAS, the leader in business analytics, is releasing an enhanced version of SAS® Text Miner. The new version features a patent-pending rules generator that helps organizations crunch through text and other unstructured data to actively learn from and discover meaning in their data.
"Active learning is a distinct advantage for SAS Text Analytics customers. By speeding up the process, they can get a quick understanding with a less labor-intensive approach," said Sue Feldman, Vice President, Search and Discovery Technologies at IDC. "Big data technologies have begun a revolution in how we understand and use information. New approaches like the enhanced SAS Text Miner help integrate the 'what' of data with the 'why' of text. Furthermore, these technologies can finally exploit the massive stores of untapped content and scattered data, turning information overload into an information advantage."
SAS® Text Miner replaces costly manual processes, improves accuracy
Sources such as blogs, newsfeeds and call centers bring in more textual data than ever before. SAS Text Miner's new Text Rule Builder node uses predictive and descriptive modeling to automatically generate Boolean rules to better classify documents. These readily interpretable rules can be directly used within SAS Enterprise Content Categorization to classify content in real time. Subject-matter experts can refine machine-generated classification algorithms, resulting in lower costs and greater accuracy.
"Many companies haven’t begun to benefit from valuable enterprise text data," said Fiona McNeill, Global Product Marketing Manager for SAS Text Analytics. "Most know that information in-house and in social media must be analyzed to bring value. SAS Text Analytics are being used for patient safety in health care, digital content performance in the media industry, early-warning systems and citizen intelligence in government and more. Nobody delivers the depth and breadth of technology for analyzing structured and unstructured data that SAS does."
In addition to many other enhancements, the new release also offers a new markup matcher that simplifies acquiring Web data for research.
Discovering knowledge in text documents
SAS Text Analytics is helping organizations uncover critical insights they can act on by applying linguistic rules and statistical methods to automatically assess and analyze electronic text in social media content, call center logs, survey data, emails, loan applications, service notes, insurance or warranty claims and more.
"SAS Text Miner helped us analyze detailed call center notes, and we achieved a much more granular understanding of callers than we had before," said Greg Hayworth, Scientist for Humana's Provider Network Operations. "Other projects include comparing the authorizations from nurses who review medical claims to the medical services that were provided and, another, analyzing free responses in surveys of our health care providers. Reading each note was inefficient. SAS helps us fully benefit from this information."
Boosting results with SAS® Text Analytics
In addition to SAS Text Miner, other modules within the SAS Text Analytics family include:
SAS Enterprise Content Categorization drives faster, more relevant information organization, search and retrieval. Add-on modules offer prebuild taxonomies, Web and file system crawling, faceted search, duplicate document identification, text summarization, and more.
SAS Ontology Management maximizes text repositories by linking them consistently and systematically using semantic relationships.
SAS Sentiment Analysis locates and analyzes sentiment in real time in electronic text from websites, communications and internal files to spot trends and identify customer priorities.
Today's announcement came at the Analytics 2012 event in Las Vegas. Presented by SAS, the conference brings together more than 1000 analytics experts and practitioners.