Thursday, February 10, 2011

Step 3. Organize—Process Template


The organization step captures information about the document (such as title, parties, and jurisdiction), produces an outline of the document, and creates clause libraries for each section of the outline. All these steps can be performed manually, but they are much faster and more accurate when processed with technology.

Task: kiiac consultant processes the document collection, creates a template outline, and clause libraries.

Index documents:
  • Classify document type
  • Extract metadata (e.g. title, parties, law firms and jurisdictions)
  • Decompose into clauses
Generate template:
  • Create a template outline
  • Clause libraries
  • Generate review reports

(a) Metadata

Extract metadata from the source documents such as: title, parties, date, law firms and jurisdiction.

(b) Outline

Review all the documents and identify the common organizational theme. In some instances, the documents may be organized in a consistent manner. For example, all documents may have a section titled Miscellaneous or General Provisions. In other situations, similar clauses may appear in different sections of the outline. The program or reviewer must determine whether:
  • clause sections are distinct (such the organization clause that may appear in both the representations of the buyer and the seller);
  • clause sections are the same, but in different places in different documents (such as a further assurances clause that may appear in the transaction section, the covenants and the Miscellaneous provisions);
  • clause sections are overlapping (for example an amendment and waiver clause overlaps with both the separate definition of amendment and waiver clauses).
This step can be performed manually or automatically. Those who have attempted to organize a template manually will appreciate the complexity of the challenge as each organizational adjustment will have significant collateral effects. It truly is a job better performed by a computer, because technology can analyze millions of comparisons per second to find the best fit. The process is similar to the board game Mastermind; but played by a computer, which can solve the code in a split second.

(c) Clause Library
For each branch in the outline identify all the matching clauses. Within these clauses find the core language, namely, the clause that contains all the common terms and the less amount of divergent or deal-specific language. In addition, identify the full range of language variations.

See, How Does it Work for more information on the computerized process.

No comments:

Post a Comment