TAMI I for Interrogative Domains

TAMI I performs content analysis on digital interrogative transcripts such as trial transcripts, surveys, hearings, interviews, and interrogations. TAMI I discovers new, previously unknown information and knowledge automatically from the content of the interrogative transcript and can then answers who, what, where, and when questions about the objects and key concepts entailed within the interrogative domain transcripts.

  1. Transcript Browsing:

TAMI I allows the user to view the transcript as HTML, PDF or plain text document. The user can use a keyword search to quickly find any word or phrase in the transcript. All browsing is done with an easy to use graphical browser interface.

  1. Analysis:

TAMI I provides a statistical analysis of many of the items and objects found in the transcripts. Questions and answers can be categorized by type and frequency and objections can be viewed per attorney or by rulings. The relationships between Q&A pairs are analyzed according to their function and summarized.

  1. Mining:

The Q & A sets are evaluated and entailed propositions are discovered. Statements that can be inferred from the entailed propositions are identified. Semantic and pragmatic relationships between statements are uncovered.


MR. GOLDBERG: Okay. And other than that, do you have any other specific recollections in terms of the types of shoes?

MR. Poser: Umm, he bought--he liked to buy dress casual shoes, not shoes to wear with jeans and so on.

MR. GOLDBERG: Even though--well, did you recognize that he had some type of celebrity status when you first started selling to him?

MR. Poser: Yes, I did.

MR. GOLDBERG: Even though he had some type of celebrity status, did you--considering that he had that status, did you somehow also memorize the occasions when you sold him shoes and what you sold him?

MR. Poser: No.

MR. GOLDBERG: All right. Now, during the time period that you had sold shoes to Mr. Simpson, was Bloomingdales one of the forty stores that sold Bruno Magli Lorenzo shoes?


(Transcript Analysis Mining Interpretation Software System)

We perform content analysis on digital transcripts of structured and semi-structured documents such as trial transcripts, surveys, hearings, interviews, interrogations, emails, and system logs (Epistemic Logging). By using text and data mining, concept/entity identification, sophisticated natural language processing, and knowledge representation methods, our TAMI software systems extract propositional knowledge, relationships and classifications then creates high quality visual summaries and visualizations.

  1. Interpretation:

Total transcript consistency is

reported. Opening/closing

arguments are evaluated against testimony given  for  logical consistency.  The verdict is also evaluated (from a purely logical view) relative to testimony given.