Company Dossier Product Emerges from LEXIS-NEXIS’ SmartIndexing Technology

November 8, 1999 — The NEXIS unit of LEXIS-NEXIS (http://www.lexis-nexis.com) has developed a new product called Company Dossier, based on its advanced SmartIndexing technology. Available through the LEXIS-NEXIS Universe Web service, the Company Dossiers draw information from all over LEXIS-NEXIS' vast array of business and legal data, pulling the information into topically arranged reports.

Users enter the names or ticker symbols of companies. They can also create lists of available Company Dossiers using criteria such as industry, assets, liabilities, sales and revenue, earnings per share, country of headquarters, etc. Company Dossiers are available for U.S. public and private companies, as well as non-U.S. companies. Though the list of companies covered changes, the company has begun with all U.S. public companies traded on major stock exchanges and a large selection of prominent private companies. NEXIS plans to add a lot more European companies next year.

Each Company Dossier displays up to eight topical reports per company, depending on whether the system has information available for each category. The eight topical reports include the following:

Snapshot—"At a glance" analysis of current information about a company, including news articles and press releases
In The News—By top publication
In The News—By topic
In The News—By region
Business Analysis—General business information including competitors, products, and brand listings
Financial Analysis—Current and historical financial information
Legal Analysis (coming soon)—Current legal information including court decisions and jury verdicts
Intellectual Property Analysis (coming soon)—Current intellectual property information, including patents, trademarks, and copyrights

Beyond all the news sources from which the system draws, Company Dossier also draws on a number of other sources, particularly for private companies, that include America's Corporate Finance Directory, Directory of Corporate Affiliations, Hoover's Company Capsules, Hoover's Company Profiles, Political Action Committee reports, Mergerstat M&A reports, and Investext reports. Where appropriate, Company Dossier will also link to company listings in infoUSA US Business Directory, Gale Group Company Intelligence Report, Extel Annual Card, Executive & Directory Biographies, and Standard & Poor's Register of Corporations.

Each report presents extracted information in tabular format for easy downloading into spreadsheets. Reports also link to the full-text documents supplying the presented information, including the source of the information and the date the source was last updated. Hotlinks in the reports connect to company e-mail addresses and Web URLs. Using the Custom Dossier option, users can select specific sections of each report for packaging into a single report.

Company Dossiers will appear on LEXIS-NEXIS Universe in December. LEXIS-NEXIS demonstrated the product at the Online World Conference and Expo in Chicago, October 25-27.

As of this writing, LEXIS-NEXIS had not issued price information for Company Dossiers, beyond indicating that it would be subscription-only as an add-on to some existing packaged content. No pay-as-you-go access is permitted.

SmartIndexing Technology
LEXIS-NEXIS has followed a policy of supplying additional invisible indexing to records supplied by its long list of information providers all the way back to its Mead Data Central days. However, the new SmartIndexing technology has allowed a substantial expansion of this practice. LEXIS-NEXIS has used automated systems to help index proper names since 1992, tagging "articles discussing large public and private U.S. and international companies, newsworthy individuals, large organizations and government agencies, and all countries and major cities worldwide."

In 1998 the company expanded the process into topical indexing, which currently covers over 900 subject categories with more coming. Throughout 1998 and 1999, LEXIS-NEXIS has expanded the topics indexed by subject indexing the sources in the CURNWS library (most recent 2 years of news content). The expansion of controlled vocabulary terms covered several industries—media and publishing, aviation, health care, entertainment, computing and technology, banking and financial services, insurance, pharmaceuticals, and small business. The general news terminology expanded for politics, weather, social and global issues, and sports.

Currently, the SmartIndexing technology indexes news documents for the following:

30,000 company names
20,000 people names
10,000 organization names
800 geographic locations
More than 900 business, industry, and news topics

For the last category of additions, LEXIS-NEXIS uses professional librarians in collaboration with subject experts. The librarian indexers identify significant terms and concepts, and write algorithms that can analyze the actual content of documents. The human contribution includes identifying key document attributes such as headline, body, publisher-provided indexing, etc. SmartIndexing technology uses these algorithms to read, analyze, and index new documents as they flow into the system.

The automated indexing programs associate a percentage score with each indexing term on a document, which measures the amount of discussion in the document as compared with the indexing algorithm. They apply a relevance score or percentage (based on criteria such as a term's frequency, weight, document placement) for each indexing term in LEXIS-NEXIS news articles. These percentage scores are interpreted as follows:

90%-99%=Major Reference (significant discussion of the topic)
80%-89%=Strong Passing Reference (not the main focus of the article, but discussed)
50%-79%=Weak Passing Reference (some mention)

Users can search on relevance percentages, for example, retrieving only the highest-scoring documents in a precision-oriented strategy or going for recall by retrieving all documents on a topic, even passing references, but with the addition of controlled terminology that the system has assigned. To search on a topical indexing term, searchers simply do a Subject segment search, e.g., "subject(index term)". To retrieve on a major reference, use "subject(index term 9*%)"; on a major and strong passing reference, use "subject(index term PRE/2 8*% or 9*%)". (For a list of topical indexing terms and a general description of the SmartIndexing technology, visit http://ip.lexis-nexis.com/products/index.)

Below is a sample of a topical index definition developed for the concept of "joint venture."

SCOPE NOTE: JOINT VENTURES targets joint partnerships between companies where a third, new company is formed and co-owned by the partners. The
scope includes announcements of new ventures, failed ventures and joint venture banks.

THRESHOLD = 4 {minimum score needed to tag a document}

HEADLINE COUNT = 2 {score given to a phrase matching in the headline segment}

BODY COUNT = 1 {score given to a phrase matching in the body segment}

VERY STRONG TERMS
"concentrative" j/v
"concentrative" j/v
"concentrative" joint venture
"concentrative" joint ventures
approve the j/v
approve the j/v
approve the joint venture
to form joint venture
to form joint ventures

STRONG TERMS
j-venture
j-ventures
j/venture
j/ventures
joint venture
joint venturer
joint venturers
joint ventures
joint-venture
joint-venturer
joint-venturers
joint-ventures

STRONG RELATED TERMS
form a venture
form the venture
form ventures
formation of a venture
formation of the venture

WEAK TERMS
% of the arrangement
% ownership
% stake
business combination
business combinations
agreement
agreements
form joint
form jointly

NEGATIVELY WEIGHTED TERMS
high school jv
joint vision 2010
junior varsity
jv 2010
jv boys
jv girls
jv squad