Friday, September 03, 2010
 
ILE Home

Research

TALE

A Telling TALE
Research
  Login     Register
Publications Minimize

Only registered users may download our peer-reviewed, academic publications. Please click here to visit our sister organization, ALIAS Technology LLC, where you can register and download publications. You can also download case studies in which ILE research has been used, whitepapers relating to forensic linguistics, and multimedia files of Dr Chaski speaking about forensic linguistics and testifying as an expert witness.
Spotllght Minimize

ILE has research partners all over the world, scholars who share our passion for forensic linguistic research and truth in justice. We spotlight researchers here.

Samina A. Khan, MS,  teaches English at the University of Management and Technology in Lahore, Pakistan, where she is working on her PhD with Dr. Anjum Saleemi, Dr Kabal Khan and Professor Nazir Malik. She has recently completed "Urdu-English Conversational Code Switching in a Pakistani Immigrant Family: A Case Study."

Ms. Khan's interest in forensic linguistics focuses on the language of suicidal and non-suicidal poets, writing in Urdu, and whether such research can then be applied as a screening devise for depressed or suicidal students in Pakistan.

Ms. Khan is able to use ILE research tools because our computing platform is Unicode compliant, and accepts the Arabic script in which Urdu is written as easily as the Roman script of English.
Current Research Projects Minimize

FORENSIC LINGUISTIC TOOLS

All of our research projects include two questions:

  • Can the accuracy of a statistical model reach a high enough level with the limited data quantity typically available in a forensic setting to be useful in investigation and adjudication?
  • Can the linguistic analysis be fully automated or semi-automated?

These questions are not included in the research project descriptions, but they are guiding principles of all ILE research. Each of the following projects is fully automated with current research focusing on developing the appropriate databases for continued validation testing and reaching the highest possible accuracy.

GIGISM Gender Guessing: with what degree of accuracy is it possible to guess the gender of a document's author?

AGNESSM Age Estimation: with what degree of accuracy is it possible to estinmate the age of a document's author? Can the accuracy reach a high enough level with the limited data quantity typically available in a forensic setting to be useful in investigation and adjudication?

WISERSM Witness Statement Collusion: how accurately can we determine if two witnesses have actually experienced the same event, or have been coached by one who did witness the event?

PRETextSM Predator Text Assessment: how accurately can we predict when an overtly predatory sexual text will occur in a chat?

FIREPANTSSM Veracity Assessment: how accurately can we distinguish between a truthful witness statement and a less than truthful statement? how accurately can we identify less than forthcoming statements in depositions and other dialogic communications

OnGOING VALIDATION TESTING

SynAIDSM: syntax-based author identification

UniAIDESM: grapheme-based author identification

ThreatAssessSM: determines statistically is a letter is a real threat or a control document

SNARESM Suicide Note Assessment: how accurately can we determine if a text is or is not a real suicide note?

FUNDAMENTAL TEXT ANALYSIS TOOLS 

In developing forensic linguistic methods, we begin by building fundamental text analysis tools. These tools are then tweaked and developed into components of ALIAS: Automated Linguistic Identification and Assessment SystemSM, which underlie the services offered through our sister organization, ALIAS Technology LLC. Some of our fundamental Text Analysis Tools are:
 
InterTexterSM: derives n-gram analysis of each document in a set and determines the overlap of n-grams from 2 to 8 words in length
 
LexiLapSM: quickly derives several measures of vocabulary overlap between two documents, or one document and a set of documents.
 
Stripper: divides a text into content words and function words with option to strip words to base form.
 
Tagger: uses a ~30,000 word lexicon, morphological rules, and syntactic rules to tag words for part of speech.
 
Scraper: enables the user to scrape text from the web directly into the ALIAS Research Document LIbrary database.
 
E-TaggerSM: Tagger with the addition of other elements for highly ambiguous tagging.
 

 
Research Document Library Minimize

Robert_Leonard_Phd_forensic_lingustic_stylistics The ALIAS Research Document Library is a multi-corpus resource for independant question document method testing and validation performed in coordination for the advancement of forensic linguistic science at the Institute for Linguistic Evidence (ILE). For more information on method validation testing and corpus access inquiry, please contact Dr. Chaski .

Robert_Leonard_Phd_forensic_lingustic_stylistics Recent Additions to the ALIAS Research Document Library

December 2009

  • Real Threat Letters based on a FOIA request and donations

November 2009

  • False and True Trauma Narrativec based on experimental collection
ILE Mini-Grant Funding Minimize

The following announcement was posted on the Linguist-List and sent to forensic linguistic researchers in June 2009.

The Institute for Linguistic Evidence, Inc., is pleased to announce a grant competition for research, data collection, and/or validation testing in forensic linguistic methodology.
 
ILE plans to award a maximum of five mini-grants (US$500-US$2500) in summer 2009. Validation testing and data collection are the highest priorities, with preference for testing of existing systems such as ALIAS. Grantees should have earned at least the MA in linguistics, computer science or statistics, with the PhD or ABD preferred. For an application packet, please contact, via email, Carole E. Chaski, PhD, Executive Director, at cchaski@LinguisticEvidence.org.
Current ILE Mini_Grants Minimize

The 2009 ILE MIni-Grant Funding Competition Announcement yielded requests for the application and completed proposals from the United States, Europe and Africa.

Ghenba Ibileye, PhD "Ransom Notes in Nigeria"

Roger Stritmatter, PhD "Authorship Testing of the Hydrachos Satire Using ALIAS"
ILE Home | Research | TALE | A Telling TALE
Copyright 2009 by Institute for Linguistic Evidence, Inc.
Terms Of Use | Privacy Statement