SentiMap - Sentence Handling Toolkit : exploit English sentences in your business applications

Sentence Analyzer - enabling business and enterprise applications to handle sentences and text

Executive Profiles in company filings How much time would it take to scan and obtain a desired match, from among 100,000 U.S senior executive profiles ?

This demo is about picking data from profiles and resumes. This kind of text is often less structured than formal financial statements.

Scenario : The below kind of executive profiles are found in Annual Reports and Definitive Proxy (DEF-14) filings of companies. The high resolution document extractor at text2data.net will extract most tables and structured data , even though DEF-14s yeild less easily to such IR techniques.

Where a rigid structural extraction process can no longer provide confidence in the extracted data, this sentence analyzer can step in and fill in the remaining.

Try out the below , and sometimes, explore the "Compare less tightly" and other options. For questions : kinshuk_in @ yahoo dot. com

Note : The RESULTS shown are the human readable equivalent of Java/C# objects, and they have lots of additional information, like word meanings, group codes like colors and flavors etc., intended for further analytical/statistical treatment. This demo (created with no NLP APIs), stresses that with text, it is better to first maximize grammar based processing, and use statistics/math methods much later.

Sentences must be separated from each other by an ending period (. or ! or ?) and one space.

Skip the descriptive stuff and go directly to demo

Executive Profiles : finding data ( Click more samples : then Find/Collect)

Enter DESIRED benchmark sentence(s), or click one of the sample buttons above.. Max. 5 sentences.

Enter TARGETed sentence(s) or a paragraph. [See an Exec Profile extract below]. This is what you want compared against the benchmark(s). Max. 20 sentences.

Bevin Desmond is president of Morningstar's international division, responsible for identifying and developing new markets, managing and directing operations, and launching new products. She is also president of institutional software, overseeing Morningstar's products for institutions including Morningstar Direct, a Web-based institutional research platform. Desmond joined Morningstar in 1993 and was one of three employees who started the company's international business in the late 1990s. From 1999 to 2000, she served as manager of all international ventures. From 2000 to 2008, Desmond was president of Morningstar's international business. Desmond holds a bachelor's degree in psychology from St. Mary's College. David Williams is a managing director of Morningstar, Inc., responsible for corporate design and its application to brand identity, products, communications, and the workplace. Williams joined Morningstar in 1993 and has been instrumental in establishing design as one of the company's core capabilities. Under his leadership, Morningstar has been nationally recognized for design quality in product marketing and communications. Williams holds a bachelor's degree in industrial design from Ohio State University and a master's degree in fine arts from Yale University School of Art. He is a member of the American Institute of Graphic Arts. Chris Boruff is president of advisor software for Morningstar, overseeing strategy, development, and marketing associated with Morningstar's products for financial advisors including Morningstar® Advisor Workstation and Principia. Boruff joined Morningstar in 1996 as product manager for Principia, the firm's CD-ROM-based investment research and planning software for advisors. From 1997 to 2000 he served as senior product manager and head of Advisor products. From 2000 to 2008, he was president of the Advisor business unit. He holds a bachelor's degree in economics and psychology from Northwestern University.

(Please scroll down for the RESULTS)

Summary results : Searched and sorted among 15 target sentences. Highest co-relation 120. percent.

Best matching content + structure after a find/collect operation

All sorted matches (Best finds on top, degrades towards the end, and very bad matches ignored)

Options for comparing, or finding and selecting (i.e. change default settings)
Compare less tightly	Use frequently, if find results are not OK
More structural than content	and
Tree depth (deeper into structure)