Product • SeekiNF

SeekiNFFeatures & Technology

SeekiNF search engine empowers users to fetch financial and non-financial information by querying the system and perform various kinds of textual analysis. It does not have stop words like "in", "not", "to", "we", etc. in the search process—enabling searches such as "We do not have an audit committee financial expert serving on our Board of Directors" or "not effective internal controls" in 10-K filings.

Overview Search & Retrieval Document Metrics Similarity Measures Readability & Risk AI Tools

Overview

With its built-in features, SeekiNF allows users to obtain any financial or non-financial information from corporate regulatory filings by simply querying the system without any knowledge of programming. In addition, SeekiNF has built-in features that allows users to:

  • perform various kinds of textual analysis such as word count, phrase count, and sentence count.
  • obtain seven readability indices, and seven risk metrics (financial risk, litigation risk, tax risk, etc.).
  • determine the Cosine Measure of Similarity and Vector Measure of Similarity between two documents of the same company or two different companies.

Document Metrics

  • Word count, Word count without numbers, Sentence count available for all documents.
  • Proximity counter with Order and without Order, available for all documents.
  • Word Distribution, available for all documents.
  • Tone Dispersion, and Word Variation, year-to-year, all documents.

Similarity Measures

  • Cosine Measures of Similarity—three kinds: all words including numbers; all words without numbers; words without numbers and stop words—within company and for one company versus another—available for all the SEC filings, PCAOB Reports and Indian annual filings at SEBI. See the Users’ Guide for interesting fraud related examples. One can compute this measure using their own set of phrases/words through the custom service.
  • Vector Measures of Similarity (Cosine measure has a severe deficiency; Dr. Srivastava proposed a better measure)—three kinds: all words including numbers; all words without numbers; words without numbers and stop wordswithin company and for one company versus anotheravailable for all the SEC filings, PCAOB Reports, and Indian annual reports. One can compute this measure using their own set of phrases/words through the custom service.

Readability & Risk Metrics

  • Seven Readability Indices—available for all documents—including Modified Fog Index based on the paper by Kim, et. al in Contemporary Accounting Research, 2019.
  • Six Risk Sentiment metrics (Financial Risk, Litigation Risk, Tax Risk, Idiosyncratic Risk, System Risk, and Overall Risk) for 10-Ks, 10-Qs, 20-Fs, 40-Fs, Item 1A in 10-K, MD&A and Footnotes in 10-Ks and 10-Qs.

AI Tools

Get Highlights and Summary reports for 10-Ks (Whole, MDNA, Footnote, and Risk Factors), 10-Qs, 8-Ks, and Def14As reports using OpenAI.