PsychoThesaurus

This is a project related to psycholinguistics and psychology.
…

…

PsychoThesaurus
Industry: AI, Digital Health Care
Problem:  Common mental disorders lead to considerable losses in health and functioning.
Depression is a common illness worldwide, with >  264 million people (4% men and 8% women) affected (GBD 2017) . Depression is a leading cause of disability worldwide (WHO 2020)
The proportion of population with anxiety is estimated to be 3.6% (WHO 2015)
788 000 people die due to suicide (WHO 2015)

SOLUTION
Preamble
A person describes reality with the help of language which does not belong to them. But language allows us to express ourselves through intonation, choice of words and genres we use. Our emotions, personal traits and even mental health issues may be revealed in language.
If there is a way to identify all this through speech analysis there will be a way to monitor labour productivity and probably even to predict speaker’s behavior.

Examples of personal markers in speech

This is an innovation research, aimed at creating a product (PsychoThesaurus) that will allow to identify emotions in speech and to make conclusions about the person.
Modules:
Analyzer – finds tokens and marks them grammatically. Thesaurus – a multilevel list of tokens and collocations that fall into the following thematic groups:
areas of life (person, society, nature, physical world, etc.);
emotion (anger, fear, sadness, etc.);
emotional and semantic dominant (light, dark, merry, sad, beautiful – Belyanin 1996, 2000).
Categorizer – tokens are put into correspondence with a specific thematic group in Thesaurus. E.g. words tired, ill, dull, bored will fall in the group ‘tiredness’, and the words space, nature, navigator will fall into ‘travel’.
Diagnosis – compares the completion of the groups with the type of the text. Thus the prevalence of tokens related to ‘tiredness’ may indicate the low mood of the author. And the prevalence of words related to ‘travel’, ‘sport’, ‘money’, ‘appetite’ may indicate inclination to risk activities.
Summarizer – presents the results in text and visual form.
Market
Companies that have a need in analyzing texts in order to mine emotions and indicators of mental health issues of the authors.
Social network organizers.
Services responsible for providing mental health

Competitors

WordNet
Talkwalker
Repustate
Lexalytics
Critical Mention
Brandwatch
Social Mention
Sentiment Analyzer
ВААЛ Vaal

Business Model

Online service for companies. Per word fee / monthly fee.
Mission
Currently
·         an outline of the program is created
·         there are several dictionaries with almost 20 000 lexical units.

By the end of 2022 a minimum viable product  may be created.
By May 2023 the program may be launched in Beta-version.

The Product  may be integrated into word processors and messaging systems and will be used in social networks analysis.Team

Director – Valery Belyanin
Project manager – no
Software engineer – no
Software designer – no
Consultant – Katerina Repina (?)
Consultant – Elina Sarakaeva
Consultant – Scott Malec

Questions
If you have questions you may ask me by email .

You may also use the form: