This post is about the connection of computational linguistics and psychotherapy. You will learn how to use computational semantics as a person in search of answers and solutions in a therapeutic context. The goal of semantic association analysis is to define a manageable number of semantic standards for disorder patterns and problem descriptions that occur millions of times, which give people (therapists as well as clients) new choices.
Semantic association analysis: How to recognize and influence mental and emotional processes
Internet-based semantic analysis and Natural Language Processing NLP can provide significant services to psychology and self-intervention. Semantic association analysis is a special form of association analysis related to human psychological processes and action planning. Association analysis with reference to the human psyche was first proposed by me in April 2020.
- The main benefit of Semantic Association Analysis is to be able to change interpretations at will.
- Human experience and behavior become comprehensible and changeable – controlled by the individual who wants to change something in himself.
- Contextual conditions without a currently visible solution structure can be connected through this form of analysis with solution networks that are already present in one’s own life as experiences.
- Semantic Association Analysis is used to simplify and increase the efficiency of counseling processes such as psychotherapy, coaching and supervision.
Measurable results from computational linguistics for psychotherapy and coaching
If a person explicitly sets a goal to leave a negative state, then he has at most a neutral goal (no longer negative). Such goal formulations are predominantly associated with negative terms. This also applies to the formulation of solutions. The solution is “no longer negative” – thus contains the word and thus the effect of “negative.
I have semantically analyzed some terms from the German text corpus and translated them into English here. Above all, this is about the linguistic phenomenon of word neighborhood.
The evaluations of the following terms were done with termlabs.io and translated into English.
A person who strives to eliminate a negative condition and names his activities in this way can hardly develop a positive goal image. Everything aims at an end and not at a beginning:
- End war
- Overcome fear
- Eliminate disorder
Semantic association analysis – example “eliminate clutter”
In connection with this goal, the following adjectives occur frequently in the written language: little, daily, constantly, large – and in addition these verbs: eliminate, get along, free, sort, avoid, make, rule, clean up, push, arise, create, suffer, lead, worry.
It is similar with the spoken language. Clients figuratively describe the problems in their formulations rather than the solution view.
Example of a targeted formulation: “Keep order”.
In connection with this goal formulation the following adjectives occur: fast, effective, long-term, created, right, systematic, finally, easily, simply, zonally, decoratively, fundamentally, thoroughly, won, in addition, permanently, worldwide, beautifully – and these verbs: keep, create, preserve, maintain, sustain, disturb, view, produce, check, bring.
Even though these results refer to the written word – it looks similar in the spoken language.
People who can only imagine the end of a burdensome condition also talk about burdens in the formulation of goals:
emotional, mental, psychological, everyday, difficult, lasting, strong overcome, recognize, reduce, arise, come, get under, clear, cope, release, let go, bypass, stuck, react, get, protect, experience, fall
Those who are oriented toward relief use a different speech pattern:
Allow ease: mental, new, full, proverbial, humorous, psychic, inner, transform, allow, need, let, radiate, revive, master, transfer, train, attain, set, understand, create, come, experience, find, learn.
The language pattern used in formulating a desired change is therefore critical to the progress of the change.
Also, when describing or analyzing a problem, the language pattern provides, depending on the choice of words, either outlooks on options for action or supposed evidence for the supposed hopelessness of a situation.
Simple explanation: what are associations, what are the effects?
Human associations (from Latin (a)sociare, to unite) are voluntary or involuntary (unconscious / “automatic”) connections in the brain, e.g. between symbols (writing, language, forms), sensations (pleasure, pain) and actions or reactions (activity, passivity, withdrawal, escape, avoidance, etc.).
Examples of associations that occur without intention or volitional effort, i.e., involuntarily:
- Somatic associations (stress reactions, hormone activity)
- Mental-psychic associations (conclusions, hypotheses, extrapolations).
- In psychology, we talk about associated disorders (also: comorbidities).
Associations trigger hypotheses, extrapolations and experiences that are not reflected and lead to
Examples of non-reflective (not critically examined) associations:
- „Animals” & „dangerous”
- „Authorities” & „oppression”
- “The world” & „danger”
Unreflective hypotheses can lead to distorted perceptions and realities
Examples of unreflective projections and hypotheses:
- This animal (snake) is dangerous >>> “All animals are dangerous”
- A person with power has influenced me strongly >> “All authority is oppressive”
- I was in danger on a certain day >> “The world is threatening”
A person who has grown up in the context of violence or war will tend to be more suspicious of the world: until he has gained new experiences and integrated them into his life.
Unreflective experiences can also “unintentionally” bring about good:
A person who has been socialized in the context of safety and security will in all likelihood predominantly assume a state of security in adult life.
He latently believes himself to be safe.
Semantic Association Analysis, like Steve de Shazer’s Brief Therapy, is about looking for exceptions. If an association has already worked once in a favorable way, this effect can be repeated at will.
Willfully induced associations
Intentional associations include forms of conscious learning and behavioral self-direction
The intended (intentional) association is the result of deliberate and planned neuronal connections – and vice versa. A bidirectional effect occurs and spreads to other areas of thought.
The more networks of the brain become active, the more additional network neighbors can act – and thus, for example, derive a search for a solution with options for action from a problem situation.
By repeatedly recording and replaying information (e.g., high school graduation material) and movement sequences (playing the piano, athletics), humans use neuronal connections to create permanently retrievable memories and abilities to repeat complex sequences (e.g., playing a Bach sonata) at any given time.
The model here can be the training of networks in artificial neuronal networks – with active interest and participation of the subjects or clients.
It is a considerable plus in self-efficacy when a person realizes: What this human-built arrangement of artificial neural networks in neuroinformatics can do, my brain can do too, and even better.
Training and learning: the intention here is to have data and skills retrievable in a desired context.
Association is associated with standardization and repetition
Examples of volitionally generated associations: “I associate with …”
- Someone is leafing through a cookbook or searching for recipes on the Internet
- A person stops in front of a restaurant and reads the menu
- An overweight person thinks of food in parallel with his health and decides to drink a tea as a substitute
Willfully induced associations are effective – the desire for dissociation is paradoxical and unfulfillable
A person who wants to get rid of or control his anxiety or obsessions formulates an impossible goal.
On the other hand, if you say to yourself, “Whenever an idiotic thought, a strange thought, comes up, my brain reacts with amazement, and it quickly gets bored …”, then we are dealing here with the conscious training of a neural network. In one’s own head.
The schematic representation of a simple artificial network ANN can serve as a demonstration object.
Associations can be changed by additional associations
Even if they have already acted many times in a particular biography:
Even involuntary associations are nothing static. They can appear anywhere as if by chance. One speaks of an attitude, e.g. misanthropy in people who had bad experiences with other people.
The effect of associations can be changed by adding them.
Linking association analysis to computational linguistics.
Computational linguistics makes unstructured text in text corpora and spoken language machine-readable. It makes it possible to determine and document existing connections and orders (entities, cooccurrences and networks of information).
Of particular interest for the connection to psychotherapy is the possibility of calculating probabilities for events occurring in parallel or with a short time lag.
What does computational linguistics have to do with counseling and psychology?
In computational linguistics, the acquisition and structuring of raw data (unstructured speech in text form) serve to recognize patterns and classify individual human behavior, making it plannable.
In computational linguistics, algorithmic shopping cart analyses are used to automatically suggest suitable products for the buyer’s purchase on a website.
Here, an attempt is made to determine the probable intention of the user and to use his implicit expectation.
Computational linguistics analysis techniques can help in neuroscience, psychology and psychotherapy
Mining a text corpus forms the basis for extensive semantic analyses.
All statements published in a text corpus, e.g., of a political party delivered in parliamentary debates, enable analyses of the party’s intentions and purposes. Analyses of news websites such as Twitter provide information about sentiments and trends.
This technology, which is already established in the field of machine learning and artificial intelligence, can also be used for the benefit of people:
Brain mining will come soon.
People’s vocabulary and thought constructions: Billions of possible text corpora – and intervention options that can be standardized for this purpose.
So far, people’s expressions are captured where they appear in everyday processes:
- Tweets on Twitter
- Comments, questions and answers on the Internet, e.g. on Quora
- Chat histories in service chats
Speech recognition – Capture and interpretation of spoken text
Systems for automated speech recognition and interpretation in machine learning serve marketing and the control of infrastructural projects as well as politics.
What works for public life is equally suitable for the life of the individual.
Semantic association levels of the individual
The active vocabulary of a person: How does a person express his inner life, interests and intentions?
The thought constructions and processes: What inner images and conclusions does a person experience – and with what effects (feelings)?
Semantic levels of association between individuals, in teams, organizations and societies.
Let’s look at the individual and his self-interpretation in a system:
- National affiliation – me and my country
- Member of an organization – me and my club
- Social status – family: me and my children
- Social status – profession and assets: my house, my yacht …
Visualization of speech patterns and thought patterns provides healthy distance
Experience from hundreds of counseling sessions and from the literature specifically on hynpotherapy shows that clients who look at their speech and thought patterns from the outside can reevaluate and change them.
Problems are semantic networks of their own. They can be changed.
In many life situations, decades-old core beliefs run along like a subtitle in life:
- toxic statements
- harmful thought patterns
- thought circles
Semantic-Psychological Predictive-Analysis and Sentiment Analysis
Very few people seem to have any idea what networks they activate in their brains.
There is a great didactic need here.
With essential information about the brain, many people can say goodbye to problems from their lives just by changing their speech patterns and solution ideas.
Every sentence, every thought is in itself an autohypnotic act.
The connection to computer linguistic for psychotherapists, especially hypnotherapists
There are pictures for all linguistic terms. In a picture dictionary, we find the drawing or the photo of a boat next to the term boat. In the human psyche there is no such order and assignment.
Every word and every other signal (optical, acoustic, tactile, olfactory and gustatory) is associated with different associations in every human being.
Therefore, in terms of basic autohypnotic care, it is helpful to report typical, frequently occurring associations.
People who talk and think about X often have an inner picture of it like Y. Those who have such a picture can develop feelings of ABC about it.
Individuals who try to achieve something by avoiding a negative outcome have a significantly higher effort than individuals with a clear positive inner picture.
For Computational Linguists: Psychotherapy with Python and Gephi
Almost all concepts from computational linguistics can be translated into psychological models and applied in counseling:
- Closeness centrality – how close is one term to another in a language pattern? Example: success <–> doubt
- Nodes and Edges – at which points and around which collections of terms do problem foci form and stabilize?
- Cooccurrences – which further thoughts and related feelings are in many cases connected with a language pattern?
- Link anchor texts on the web and in everyday speech – which meanings become apparent early on in a dialog on the way to a topic change, and how do they work?
There are already a number of approaches to this and models empirically tested in psychological practice.
- What are the further implications and consequences of combining artificial and natural intelligence?
- How can the insights from machine learning through didactic re-engineering also serve individuals in their personal development?
Imagine … parts of a personal problem as a vector … and a person changes only a little …
Write to me and join me and others in my private LinkedIn group. I will send you an invitation. firstname.lastname@example.org