How Content Analyst's CAAT Provides Value
Unstructured documents and emails compose some 85% of most stored information. Whether you call it ESI (Electronic Stored Information), open source, or big data, gaining insights from it or just organizing it is a big problem for most software solutions.
This is where Content Analyst makes the difference. Designed to rapidly uncover insights from those unstructured documents and emails, our CAAT platform provides unique value to partners who integrate its capabilities into their software applications.
Unsupervised Classification
Without knowing anything about the data, CAAT will dynamically cluster all of the documents into a hierarchy of conceptually similar groups. From legal review to intelligence analysis this capability can help people focus on the important topics quickly with no up-front work and no pre-defined taxonomy to limit insights from relationships the data can show you.
Supervised Classification
When you are sorting or tagging documents that are related to specific categories, supervised classification, or categorization, is the way to go. CAAT creates a conceptual understanding of the category from example documents for a more human-like review of the documents being processed. Our approach is extremely fast, highly precise, language independent and easy to train.
Concept Searching for Relevant Information
Different people use the same terminology to describe a particular concept or idea less than 20% of the time. Concept searching allows them to find relevant information even when the text of their query differs from the text of the documents in the index. Based on LSI, Content Analyst's patented technology searches using actual concepts, not simply "stab-in-the-dark" keywords.
Optimizing Email Review and Analysis
From email threading to near-duplicate document analysis, CAAT's suite of tools optimizes the review of emails to speed up the process of focusing in on the important items. When used in combination with our categorization capabilities, these tools truly enhances a reviewer's ability to more quickly isolate relevant email-based material.
Searching Across Languages
Because it is mathematically based, CAAT is naturally and uniquely language agnostic—it can operate in any language. It supports multi- and cross-lingual searches, without prior translation, and can identify languages.
This enables two approaches to working with multi-lingual data:
- CAAT can separate collections or streams of documents into individual language indexes, giving experts faster access for review and translation of hot documents into other languages.
- Given appropriate sample sets, CAAT can build cross-lingual indices, enabling it to conceptually match text in different languages without needing prior translation of the subject documents. This can streamline an organization's ability to construct a more-complete picture of a given area of interest.
In both cases, CAAT's concept searching capabilities can also be employed to prioritize the content in order of relevance.
Identifying Similar Words and Terms
CAAT understands the conceptual similarity or relevance between every word in every document across all the documents or information in an index. Unlike keyword expansion or other techniques, each additional term that CAAT identifies using its Instant Context capability has some conceptual relevance to the word or terms being queried. This added intelligence drives better investigation of information collections.
Creating Document Summaries
Using conceptual analysis, and at the simple press of a key, CAAT creates summaries of long documents based on actual sentences within those documents. This can speed up review processes. For example, a quick, verbatim summary can help reviewers ascertain whether the document needs further investigation.
