Each ALIAS module accesses the document database, analyzes the document extracting and counting specific linguistic patterns, implements a statistical analysis of the pattern counts, and reports an answer.

Some ALIAS modules are totally automated and can be accessed through web_ALIAS, by trained users.  Trainees must meet strict qualifications and pass examinations to become users. Other ALIAS modules are semi-automated and require a trained and degreed linguist to monitor automated pattern checking and thus can only be used in ALIAS Technology's Linguistic Evidence Laboratory.

Forensic Linguistics

ALIAS provides forensic linguistic consulting to law enforcement agencies, prosecutors and defense attorneys in criminal investigations and cases, and plaintiff and defense attorneys in civil investigations and cases, as well as security consultants and human resource officers.  Our analyses have been used in depositions, trial testimony, settlement negotiations in criminal, civil and military cases. Specifically, ALIAS provides

  1. expert, quantitative analysis of documentary data

  2. peer review of any linguistic analyses for forensic use

  3. consulting expert advice

  4. expert testimony with track record of full admissibility   

   Authorship Identification

SynAID Who wrote it? Did the person who sign it actually author it? SynAIDsm is the patent-pending, syntax-based author identification method that has been repeatedly admitted as expert scientific testimony to the court room after Daubert and Frye hearings, i.e.,  under both scientific reliability and general acceptance standards for admissibility. SynAID classifies documents to authors, and predicts the authorship of an unknown document based on the known-author classification. It can be used for a pool of suspect authors and verification.

UniAIDE Which writer is the most likely author of a particular blog, letter, chat or email from within a group of potential, and possibly anonymous, individuals? UniAIDEsm is a grapheme-based author identification estimator that has been tested to rank the correct author from within a group of authors in the first five slots 90% of the time, on short documents, without a length bias.

   Text Typing

ThreatAssess Is this letter a real threat, or not? ThreatAssesssm has been tested and used in real-life scenarios to distinguish real threats from simulated threats and other control documents such as complaints and love letters. ThreatAssess has achieved a minimum of 90% accuracy in validation testing, runs very quickly, and provides an objective, statistically-based tool for your in-house intelligence analysts.

SNARE Is this note, blog or email is a real suicide note, or not? SNAREsm is the suicide note analysis software developed by ILE where testing has shown accuracy rates higher than the reported subjective analysis by psychiatrists and psychologists, and higher than other software as reported in the scientific literature, with SNARE's most recent validation testing achieving 88.6% accuracy. SNARE works especially well with very short notes, and the majority of suicide notes are indeed typically brief.

VerAssess 1 and 2 Is this statement truthful or is there some problem with the veracity? Two versions have been prototyped; a dialogic version which can be used for depositions and interview transcripts and a statement version which can be used for witness statements and other narrative text including corporate reports. Firepants 1 --the statement version-- is undergoing validation testing and is scheduled for release in the last quarter of 2010.

PREText At which point in a chat do the messages begin to lead to overt sexual invitations and become predator text? PREText analyzes chats so that law enforcement officers can strategize their covert communications. It is scheduled for release in the last quarter of 2010.

   Linguistic Profiling Assessment

Gender Guessing is a difficult but potentially very useful investigative task. ILE research is developing LPA Gender Guessing as an investigative tool for law enforcement and security consultants, with a fully-automated (web-accessible) version expected for release in late 2010. Data Collection is underway and pilot tests have been run.

When an investigator is trying to narrow a field of suspects, age estimation can be particularly useful, but it is a very difficult task. 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? At ILE, we are developing LPA Age Estimator, an age estimator on business documents and on blog posts.

Native and non-native speakers of any language can be distinguished because the non-native speaker's previous language experience and internalized grammar conflicts with the grammar of the second language. These conflicts are predictable based on the first and second languages. LPA L2 Indicator provides a report of linguistic indications that the author is a non-native speaker of English.

ILE research is also underway on American English dialect assessment and educational level assessment.

   InterTextuality

WISER Have two witnesses actually experienced the same event, or have they been coached by one who did witness the event? ILE research is developing WISER to answer this question using a database of accident reports with 10,000 documents, linguistic analysis and statistical modeling for the most accurate classification possible. WISER will be fully automated and available for laboratory use and onlline access in early 2010.

InterTexter How close is too close for comfort, when documents are purported to be totally independent of each other? InterTexter provides an answer through n-gram analysis and statistical distance measures. InterTexter helped a restaurant chain protect its intellectual property and trade secrets by showing that a former employee had indeed copied from the chain's manuals and recipes.

LexiLap How many words from one text also occur in another text?

Chars 1 and Chars 2 What are the distributions of graphic characters in a text (Chars 1)? This can be useful for determining the relatoinship between multiple texts manually. What kind of overlaps between texts occur at the graphic character level (Chars 2)? This is a fully automated solution for the manual comparison.

NOverLap What are the differences between two texts?

Computational and General Linguistics

ALIAS provides linguistic and computational linguistic consulting to major corporations, small businesses and non-profits on issues involving

  1. forensic validation testing

  2. text analysis for security-sensitive materials

  3. ontology development for security-sensitive materials

  4. medical diagnosis and doctor-patient interaction

  5. literary investigations of authorship

  6. literacy campaigns in manufacturing environments

  7. communicating literacy goals

Consulting: General, Computational and Forensic Linguistic Analysis


ALIAS Technology LLC provides training in several aspects of forensic linguistic analysis, including


Language As Evidence: Using ALIAS  (Due to the sensitive nature of the documentary data involved in forensic linguistics, this training is restricted to law enforcement, licensed private investigators, forensic digital examiners, corporate human resource officers, executive security officers and other forensic linguists). This is a two-day training.

Forensic Linguistics: Voodoo or Valid?

Validation Testing Procedures and Pitfalls

Training is typically attached to conferences or located in corporate settings. For more information and to schedule training, please contact cchaski@aliastechnology.com.

ALIAS Modules: Software for Forensic Linguistic Analysis

Training: Forensic Linguistic Analysis