Our Technology

Powering the Most Accurate Search and Classification Solutions Available

At H5, we believe that document search and classification challenges can be solved by capitalizing on the strengths of humans and technology, while mitigating the weaknesses of both. While H5 has some of the most advanced search and document analysis technology available today, we believe technology is an enabler that must be harnessed and used appropriately to meet sometimes extraordinarily complex legal objectives.

Advanced Technology, Specialized Expertise, Rigorous Processes

H5's highly accurate document assessment solutions utilize a range of technologies in concert with specialized expertise and rigorous human-executed processes tailored to the specific situation. To illustrate the importance of this combination in addressing search needs, consider the constituent processes that benefit from a combination of technology and human expertise:

  • Relevance Definition: To find something you first need to define what you're looking for. Defining relevance is a complex task. H5's analysts drive the relevance definition process, but do so informed by technology-assisted data analysis.
  • Iterative Refinement: The process of finding relevant documents is one of trial, error, and constant refinement. H5's approach employs expert-driven iterative development and refinement of instructions (e.g. search queries) provided to and utilized by H5's advanced search technology.
  • Quality Assurance: Knowing you've found what you sought cannot be automated. H5 utilizes real-time quality assurance processes and expert evaluation of information provided by analytic tools and methods.

H5's professionals possess expertise in a wide range of relevant disciplines, including linguistics, statistics, computer science, law, research, information technology, process engineering, and e-discovery. Each H5 project is staffed with qualified professionals who employ H5's processes and advanced methods to address a wide range of search needs.

Proprietary Information Retrieval, Data Mining, and Validation Software

The technology H5's professionals utilize incorporates proprietary information retrieval, data mining, and validation software. Originally developed and continually refined by H5's in-house team of Ph.D. computer scientists, linguists, computational linguists, statisticians, and other information retrieval experts, H5's technology has been on the cutting edge of information retrieval for the past decade.

The core technological components include:

  • Rapid Analysis and Modeling Module (RAMM): RAMM draws upon a constellation of embedded proprietary information retrieval, data mining, and performance measurement technologies deployed on the target data population to identify, validate, organize, and compile up to multiple millions of highly effective search queries that span all dimensions of the subject matter domain.
  • Automated Intelligence Module (AIM): AIM automates human decisions about documents. First deployed by H5 for the U.S. intelligence community, AIM applies the search models to the target data population to selectively identify the relevant documents.
  • Library of Linguistic Models (LLM): Over the past decade, H5 has developed a library containing millions of highly effective linguistic models across a vast array of subject matter domains. Developed using H5's RAMM, the library of search models can be used and customized to retrieve information with similar subject matter from any data population.