Representing thought trajectory through semantic space as visible geometric features that are updatable and responsive

Abstract

The present application provides for apparatuses, systems, and methods that can leverage the semantic embedding spaces of large language models. In example embodiments, a person's thought trajectory through semantic space can be represented as visible, namable geometric features. As a person engages in a thought process, the person can provide verbal input that can be analyzed to map the trajectory of the person's progression through semantic space and calculate and visually render the geometric features of the person's thought progression. Each idea within the thought process is a point along this trajectory. Ideas (points along the thought trajectory) can be mapped as the locations of word embeddings or sentence embeddings within a semantic space.

Description

REPRESENTING THOUGHT TRAJECTORY THROUGH SEMANTIC SPACE AS VISIBLE GEOMETRIC FEATURES THAT ARE UPDATABLE AND RESPONSIVESTATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0001] This invention was made with government support under DRL-1920682, awarded by the National Science Foundation. The Government has certain rights in the invention.NOTICE OF MATERIAL SUBJECT TO COPYRIGHT PROTECTION

[0002] A portion of the material in this patent document may be subject to copyright protection under the copyright laws of the United States and of other countries. The owner of the copyright rights has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the United States Patent and Trademark Office publicly available file or records, but otherwise reserves all copyright rights whatsoever. The copyright owner does not hereby waive any of its rights to have this patent document maintained in secrecy, including without limitation its rights pursuant to 37 C.F.R. § 1.14.BACKGROUND

[0003] 1. Technical Field

[0004] The technology of this disclosure pertains generally to data processing, and more particularly, represents a person’s thought trajectory progression through semantic space and calculates visually rendering the geometric features of the thought trajectory progression.

[0005] 2. Background Discussion

[0006] Different people think differently. Individual differences in how humans approach problems or generate new ideas are widely acknowledged in both psychometric research and in popular psychology. A substantial body of evidence across multiple fields of human psychologyindicates that these individual differences in thinking are relatively stable (meaning that a person’s way of thinking is a consistent trait of that person), but also that many aspects of cognitive style and ability can be modulated with training (e.g., Kleinmintz, 2017). Because individual differences in cognition are strongly associated with future outcomes of academic and professional success as well as personal well-being, understanding how a person thinks can have substantial value for educational institutions, employers, and individuals themselves. Though many psychometric assessments have been developed to index specific elements of thinking ability and thinking style, these assessments generally yield summative numeric scores that are limited in their utility and interpretability. Current solutions generally involve numeric scoring of outcomes of thought processes (rather than capturing the thought processes themselves). Examples include cognitive, scholastic, and personality testing. These solutions also cannot enable individuals to see and interact with their thought process in order to develop / change it toward desired goals (e.g., improving creativity). A persistent challenge has been to provide more than simply a number that reflects the result of a thought process, but instead to metrically capture the thought process itself.BRIEF DESCRIPTION

[0007] To address this challenge, exemplary (example) embodiments, which can be one or more exemplary embodiments, in accordance with the present application, referred to herein as the “Shape of Thought” system (and also as “Shape of Thought,” “SoT,” “SoT system, or “the system”), leverages the semantic embedding spaces of large language models (“semantic spaces”). The Shape of Thought system provides a developed metric that can represent a person’s thought trajectory through semantic space as visible, namable geometric features. Thought can comprise, for example, a creative performance, cognitive performance, idea progression, idea generation, and the like (e.g., as they think about a problem or generate novel ideas), as a person’s thought process happens, the person(also referred to herein as “participant” or “subject”) can provide verbal input that reflects or is representative of that thought process, to the SoT system. The verbal input can be received by the SoT system via, for example, a microphone associated with the SoT system, a touchpad, a keyboard, a stylus, or any other device that can be used to provide input. Based on the verbal input, and an analysis of the verbal input, the Shape-of-Thought system maps the trajectory of the progression through semantic space and calculates (and visually renders) the geometric features of this. Each idea within their thought process is a point along this trajectory. Ideas (points along the thought trajectory) can be mapped as the locations of word embeddings, sentence embeddings, or topic embeddings within a semantic space. For example, if a person is generating different solutions to a transportation problem involving moving a house, they might first think about physically transporting the whole house. Then they might consider 3- D printing the house in a different location. Then they might consider deconstructing the house and moving it in parts that can be recombined. Each of those three ideas has a mappable location in semantic space (i.e. , the location of the sentence embeddings for each of those three sentences). Many features of the trajectory of the thought process through those ideas can be calculated and visually represented. The Shape-of- Thought system can calculate in high-dimensional spaces in which large language models are constructed, but the Shape of Thought system also provides for methods of dimension reduction that allow the Shape-of- Thought system to operate in three and / or two dimensions, such that shapes (and shape features) can be readily visualized for both measurement and interactive applications. Capturing a thought process in a way that is intuitively visually understandable to a human reader might have a wide range of potential benefits, including (but not limited to):

[0008] 1 . Employers and universities identifying individual differences in thought processes (not just thinking outcomes) as a means of selecting candidates, assigning responsibilities to specific individuals, etc., which can be used to predict future academic and professional success.

[0009] 2. Individuals understanding their own thought processes,

[0010] 3. Revealing the nature of changes in thought process (i.e. , seeing how the shape of thought processes changed, not just a higher or lower score),

[0011] 4. Training / developing creativity and other goal-relevant forms of thinking by showing a person the shape of their own thought process in real time (or after the fact) as a means to redirect or otherwise modulate their thinking (e.g., via gamification). Showing a person the features of their thought process can be done in a way that interactively allows them to develop their desired features, as further described below.

[0012] 5. Educators or employers understanding their students’ or employees’ thought processes and using this as a basis to tailor educational or training curricula to those students, and using changes in the shape of thought processes to evaluate the efficacy of curricula.

[0013] 6. Understanding differences in thought processes that relate to professional and / or personal compatibility (e.g., more similar or more complementary or diverse shapes of thought for collaborative work). In other words, it can be used to identify diversities and compatibilities of thinking styles (e.g., for building collaborative teams).

[0014] 7. Understanding the thought processes that different customers use to evaluate which products to buy and, relatedly, predicting the needs, preferences, and thinking styles of individual customers and potential customers (e.g., to identify a kind of marketing message that fits a customer’s style of thinking).

[0015] 8. Metrically capturing and visualizing individual’s thought processes(not just numerically scoring the products of these thought processes).

[0016] 9. Predicting what a person will prefer or enjoy.

[0017] 10. Reducing bias / disparity in selection of applicants.

[0018] 11. Assessing the efficacy of trai n i n g / cu rricul a to develop creativity and other aspects of goal-relevant thinking.

[0019] 12. Identifying the educational needs of students based on the geometric characteristics of their thought process.

[0020] 13. Providing that a person has been trained, or completed training, using the SoT system. This verification can indicate whether a person is more suited for a particular task, job, or school. As an example, the SoT system can store a data element (e.g., in memory device comprising a database or repository) indicating that a person has completed training using the SoT system. The person can indicate this completion by displaying a visual indicator such as a badge, flag, certificate, line on a resume, or some other visual indication that the person has completed SoT training. The data element can be queried to verify certification of completion of training using the SoT system.

[0021] Further aspects of the technology described herein will be brought out in the following portions of the specification, wherein the detailed description is for the purpose of disclosing preferred embodiments of the technology without placing limitations thereon.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0022] The technology described herein will be more fully understood by reference to the following drawings which are for illustrative purposes only:

[0023] FIG. 1 shows a flow chart illustrating example operations that can be performed by the SoT system, in accordance with example embodiments of the present application.

[0024] FIG. 2 depicts a visualization of a plot related to creativity and volume in which the creativity and volume are both high, in accordance with exemplary embodiments of the present application.

[0025] FIG. 3 depicts a visualization of a plot related to creativity and volume in which the creativity and volume are both mid, in accordance with exemplary embodiments of the present application.

[0026] FIG. 4 depicts a visualization of a plot related to creativity and volume in which the creativity and volume are both low, in accordance with exemplary embodiments of the present application.

[0027] FIG. 5 depicts a visualization of a plot related to creativity andFractional Anisotropy (FA) in which the creativity and FA are both high, in accordance with exemplary embodiments of the present application.

[0028] FIG. 6 depicts a visualization of a plot related to creativity and FA in which the creativity and FA are both mid, in accordance with exemplary embodiments of the present application.

[0029] FIG. 7 depicts a visualization of a plot related to creativity and FA in which the creativity and FA are both low, in accordance with exemplary embodiments of the present application.

[0030] FIG. 8 shows a rendering depicting how sentence-based semantic network can have nodes, and each node can represent each sentence, in accordance with exemplary embodiments of the present application.

[0031] FIG. 9 shows a diagram depicting an exemplary embodiment of a computing device that can facilitate the Shape of Thought methods and operations described herein, in accordance with example embodiments of the present application.DETAILED DESCRIPTION

[0032] Shape of Thought (SoT) System Features

[0033] In accordance with exemplary embodiments of the present application, the Shape of Thought system can capture the spatial representation of textual embeddings in the context of assessment (e.g., creativity assessment, problem solving assessment, etc.). This process can start by extracting word, sentence, or topic embeddings from the library of all unique words, sentences, or topics that were generated by participants (subjects) that the Shape of Thought system is evaluating. Vectors are created by extracting word, sentence, or topic-embeddings (BERT, OpenAI’s Ada2, etc.). These embeddings represent the relationship of a certain text (word, sentence, or topic) to all other words, sentences or topics using hyperdimensional vectors (also called “Semantic Space”).

[0034] The Shape of Thought system can reduce the hyperdimensionality (more than three dimensions, e.g., hundreds to thousands of dimensions) of embeddings using, for example, the Uniform Manifold Approximation andProjection (UMAP) algorithm, so that every word or sentence is represented by a two or three-dimensional vector that relates to all other vectors. Each performance can be lemmatized, segmented, and tokenized, making it distinguishable in the semantic space of the embedding library being used.

[0035] Another way to reduce the complexity of a semantic space is by segmenting it into larger segments than words and sentences. To accomplish this, two methods can be used: topic modeling, and k-means clustering.

[0036] Example operations (i.e. , methods) can be performed by the Shape of Thought system, which can comprise one or more computing devices (e.g., computer 900 as shown in FIG. 9 below) having a processor and memory (or other non-transitory computer-readable medium or storage device). The memory can store machine-readable and executable instructions that, when executed by the processor, facilitate performance of the example operations described here. The instructions can comprise one or more software modules (for example, modules for implementing features related to volume, fractional anisotropy, idea spacing, linearity, etc.). FIG. 1 depicts an example of such data processing and display operations 100, and various features and aspects of the operations 100 are elaborated upon further, after the description corresponding to FIG. 1.

[0037] At block 105, the operations 100 can comprise receiving a plurality of first verbal data inputs, wherein each of the plurality of first verbal data inputs are reflective of the subject’s thought, and wherein the first verbal data inputs comprise words or sentences.

[0038] At block 110, based on the plurality of first verbal data inputs, the operations 100 can comprise determining, using an analysis module executed by the SoT system, a plurality of data elements.

[0039] The data elements can comprise an idea spacing data element that relates to a semantic diversity measurement and a novelty of ideas measurement. The idea spacing data element can be derived from a calculation that is an average cosine distance between an embedding of the plurality of first verbal data inputs in a two or three-dimensional space.The idea spacing data element can be indicative of a distant and diverse idea. The idea spacing data element can also be indicative of a similar or conventional idea.

[0040] The data elements can also comprise a volume data element that relates to a semantic extent of richness of creative performance measurement. The volume data element can be derived from a calculation that is the minimum volume of a multi-dimensional ellipsoid that encompasses an embedding of the plurality of first verbal data inputs in the semantic space. The volume data element can also be indicative of a creative performance measurement that is semantically large and rich. The volume data element can also be indicative of the creative performance measurement that is semantically limited and narrow.

[0041] The data elements can also comprise a fractional anisotropy data element that relates to a semantic directionality and coherence of creative performance measurement. The fractional anisotropy data element can be derived from a calculation as a ratio of a largest eigenvalue to the sum of all eigenvalues of a covariance matrix of embeddings of the plurality of first data inputs in the three-dimensional space. The fractional anisotropy data element can comprise a value of “0,” which can be indicative of an isotropic semantic space that is representative of a creative performance measurement with theme and direction. The fractional anisotropy data element can also comprise a value of “1 which can be indicative of an anisotropic semantic space that is representative of the creative performance measurement that is exploratory and diverse.

[0042] The data elements can also comprise a linearity data element that relates to a semantic progression and continuity of creative performance measurement. The linearity data element can be derived from a calculation of a correlation between an order of the plurality of first data elements and the angles between the responses, derived from embeddings in a two or three-dimensional space. The linearity data element can be indicative of a creative performance that follows a linear semantic trajectory that is representative of the creative performance being logical and consistent.The linearity data element can also be indicative of a creative performance that deviates from the linear semantic trajectory and is representative of the creative performance being surprising and unexpected.

[0043] The data elements can also comprise a semantic network nodes and edges data element related to words or sentences derived from the plurality of first data inputs. The semantic network nodes and edges data element can utilize graph theory methods (e.g., clustering coefficient, modularity) to extract additional features.

[0044] The data elements can also comprise an optimal path data element indicative of a density related to a semantic space. The optimal path data element can be calculated as an average of shortest paths between all pairs of nodes.

[0045] The data elements can also comprise a topic modeling data element comprising an arrangement of words or sentences around topics or themes based on their semantic similarity. The topic modeling data element can relate to additional features such as number of nodes / topics, themes, and semantic nodes.

[0046] The data elements can also comprise a shape uniqueness data element reflective of a quantitative distinctness of a shape from other shapes. The shape uniqueness data element can be calculated by combining multiple features of a shape into a single model to compare that combination of features across multiple shapes. The shape uniqueness data element can be calculated using Artificial Intelligence (Al) based visual algorithms to identify shape uniqueness.

[0047] The operations at block 115 can comprise mapping, by the device, a trajectory based on the plurality of data elements, wherein the trajectory comprises one or more geometric features displayable on a display device. The trajectory comprising the geometric features can then be displayed. As will be further detailed below, once a subject sees his or her though in the form of geometric shapes, the subject can respond to the feedback by changing the verbal input that they submit (e.g., another set of words or sentences).

[0048] The operations at block 120 can then comprise receiving, by the SoT system, a plurality of second verbal data inputs related to this response to an evaluation of the trajectory by the subject, wherein the plurality of second verbal data inputs is representative of a second set of words or sentences. Based on the second set of verbal inputs, the SoT system can then determine another set of data elements, and map a second trajectory. The subject can continue to provide inputs until satisfied with the “shape” of his or her thought, or if the subject discontinues, the operations can end, represented by block 125. This interactive display and response, which can be considered real time or near real time, can be used to train a subject to respond in a manner consistent with a desired personality, skill, occupation, etc.

[0049] More specific features

[0050] Idea Spacing

[0051] In accordance with exemplary embodiments of the present application, in this feature, idea spacing, the Shape of Thought system measures the semantic diversity and novelty of the ideas generated by participants. Idea spacing can be represented or embodied as one or more data elements. Idea spacing represents the distances between the word- or sentence-embeddings in a three-dimensional space. It is calculated as the average of the cosine distance between all of the embeddings of all words or sentences in the two or three-dimensional space. Higher idea spacing indicates that the ideas are more distant and diverse, while lower idea spacing indicates that the ideas are more similar and conventional. Previous research has shown that idea spacing is highly predictive of human creativity ratings of creative stories.

[0052] Volume

[0053] In accordance with exemplary embodiments of the present application, the Shape of Thought system can account for volume. Volume measures the semantic extent and richness of the creative performance. Volume represents the spatial extent of a thought in its embedding space. Volume feature is the minimum volume of a multi-dimensional ellipsoid,which can be the semantic space encompassing the vectors representing the semantic behavior of creative performance. Volume can be represented or embodied as one or more data elements. It is calculated as the minimum volume of a multi-dimensional ellipsoid that encompasses the embeddings of the words or sentences in the semantic space. Higher volume indicates that the creative performance covers a larger and richer semantic space, while lower volume indicates that the creative performance is more limited and narrower. Volume is calculated using both the high-dimensional and the three-dimensional embeddings, as they may capture different aspects of the semantic space. It is also possible to calculate “Volume” in two- dimensional space, however, it is then the surface-area of the two dimensional ellipse that encompasses the embeddings of the words or sentences in the semantic space.

[0054] FIG. 2 through FIG. 4 shows different captured aspects of volume. FIG. 2 shows Creativity and Volume that are both high. FIG. 3 shows Creativity and Volume that are both mid; and FIG. 4 shows Creativity and Volume that are both low.

[0055] Fractional Anisotropy (FA)

[0056] In accordance with exemplary embodiments of the present application, the Shape of Thought system can account for Fractional Anisotropy (FA). FA measures the semantic directionality and coherence of creative performance. FA describes the directionality or anisotropy of the thought, revealing whether ideas tend to stretch predominantly in one direction, or are evenly spread in multiple directions. FA can be represented or embodied as one or more data elements. It is calculated as the ratio of the largest eigenvalue to the sum of all eigenvalues of the covariance matrix of the embeddings of the words or sentences in the two or three- dimensional space. FA ranges from 0 to 1 , where 0 indicates that the semantic space is isotropic (i.e., has no preferred direction) and 1 indicates that the semantic space is anisotropic (i.e., has a strongly preferred direction). An equation for FA can be as follows:The Shape of Thought system can relate three-dimensional semantic space, and more specifically, the ellipsoid that was calculated using the volume analysis to the water molecule shape that is calculated during tensor diffusion imaging (DTI) to calculate the directionality of water molecules. This directionality can be used to assess the directionality of the idea-space. FA can be interpreted as the shape of the ellipsoid that represents the semantic space: higher FA corresponds to a more elongated ellipsoid, while lower FA corresponds to a more spherical ellipsoid. Higher FA may indicate that the creative performance has a clear semantic theme or direction, while lower FA may indicate that the creative performance is more exploratory and diverse.

[0057] FIG. 5 thought FIG. 7 depict plots with different FAs. In FIG. 5, the Creativity is High, FA is High. In FIG. 6, the Creativity is mid, and the FA is mid. And in FIG. 7, the Creativity is low, and FA is low.

[0058] Linearity

[0059] In accordance with exemplary embodiments of the present application, the Shape of Thought system can account for linearity, which can be represented or embodied as one or more data elements. The linearity feature measures the directionality of semantic progression and continuity of the creative performance of thought progression, indicating how ideas flow from one point to the next. It is calculated as the correlation between the order of the words or sentences and their embeddings in the three-dimensional space. Higher linearity indicates that the creative performance follows a linear semantic trajectory, while lower linearity indicates that the creative performance deviates from a linear semantic trajectory. Higher linearity may indicate that the creative performance is more logical and consistent, while lower linearity may indicate that the creative performance is more surprising and unexpected. Linearity can becalculated by looking at the angles between the individual vectors (representing the embeddings of sentences or words). Linearity can be visualized by plotting the angle change as a function of the sequence of the responses, or as the distance travelled on the semantic space. Another variable that characterizes linearity is the Angle Volatility, which is the mean standard deviation of the angles created by the performance. Angle Volatility can provide an indication of how sporadic the performance is, meaning, whether there is consistent angle change or not, and whether this change is of large or small magnitude.

[0060] Simplified Semantic Network

[0061] In exemplary embodiments in accordance with the present application, the Shape of Thought system can implement this method, using data elements to implement the method. This method, a simplified semantic network method, involves creating a network of semantic nodes and edges based on the words or sentences generated by the participants. The nodes can be either the words or sentences themselves, or the clusters of words or sentences based on their semantic similarity. The edges can be either the co-occurrence or the semantic distance of the words or sentences. The simplified semantic network can be used to visualize and explore the semantic space of the creative performance, as well as to extract additional features using Graph Theory methods, such as: a. Clustering Coefficient This feature measures the degree to which the semantic nodes tend to form cliques or groups. It is calculated as the average ratio of the number of edges within a node’s neighborhood to the maximum possible number of edges. Higher clustering coefficient indicates that the semantic space is more modular and clustered, while lower clustering coefficient indicates that the semantic space is more uniform and dispersed. Clustering coefficient can be represented or embodied as one or more data elements. b. Modularity: This feature measures the extent to which the semantic network can be divided into distinct communities orsubgroups. It is calculated as the difference between the number of edges within a community and the expected number of edges if the network was random. Higher modularity indicates that the semantic space is more heterogeneous and diverse, while lower modularity indicates that the semantic space is more homogeneous and similar. Modularity can be represented or embodied as one or more data elements.FIG. 8 shows that a sentence-based semantic network can have nodes, and each node can represent each sentence.

[0062] Optimal Path

[0063] In accordance with exemplary embodiments, the Shape of Thought system can account for the optimal path, which measures the average distance between any two semantic nodes in the network. Optimal Path can be represented or embodied as one or more data elements. It is calculated as the average of the shortest paths between all pairs of nodes. A higher optimal path indicates that the semantic space is sparser and more disconnected, while a lower optimal path indicates that the semantic space is denser and more connected. a. Path Divergence (Progressive Cosine distances from the origin) measures the linearity of thought directionality, reflecting how the directionality of thought changes from the first generated word. Path divergence can be represented or embodied as one or more data elements. A linear model is constructed using the serial generation of words / sentences to predict the distances in the angle taken at each generative point compared with the previous one (word / sentence). The R2is taken as a representation of the tendency to see linear divergence during the creative process. b. Path Distance (Progressive Euclidean distance from the origin) measures the linearity of thought progression, indicating how linear is the flow of ideas from a starting point. A linear model is constructed using the serial generation of words / sentences to predict the Euclidean distances taken at each generative pointcompared with the previous one (word / sentence). The R2is taken as a representation of the tendency to see linear progression during the creative process. Path distance can be represented or embodied as one or more data elements.

[0064] Topic Modeling

[0065] In accordance with exemplary embodiments of the present application, the Shape of Thought system can implement the method of topic modeling. This method presents a visualization of the embeddings based on the content of a topic. Topic modeling involves identifying the main topics or themes in the creative performance, and arranging the words or sentences around those topics or themes based on their semantic similarity. These arrangements can be represented or embodied as one or more data elements. The topic modeling method allows capturing the thought process at a higher level by grouping related ideas and concepts, thereby providing deeper insights into the overall thought process, revealing the underlying connections of themes in creative performance. Relatedly, the topic modeling method offers additional features, such as: a. Number of Nodes / T opics: This feature estimates the number of clusters in the embeddings-space used by the participant. It is calculated using k-means clustering to generate an embeddings- based space from the responses of all the participants, and then using this model to predict the number of clusters produced for each of the responses provided by the participants. Number of nodes / topics can be represented or embodied as one or more data elements. A higher number of nodes / topics indicates that the semantic space is more complex and diverse, while a lower number of nodes / topics indicates that the semantic space is simpler and more similar. The number of nodes / topics can be represented or embodied as one or more data elements. b. Themes: This feature identifies the main themes or topics in the creative performance, and arranges the rest of the words or sentences around those themes or topics based on their semanticsimilarity. It is calculated using topic modeling algorithms to infer the latent topics or themes from the embeddings of the words or sentences, and then using dimensionality reduction techniques to project the embeddings onto a 2D or 3D space. The themes feature can be used to visualize and explore the semantic space of the creative performance, as well as to measure its coherence and diversity. Themes can be represented or embodied as one or more data elements. c. Semantic Nodes: This feature can also be used as semantic nodes in the above-mentioned Simplified Semantic Network. Compared to word or sentence nodes, network analysis using theme nodes can offer a more abstract level of understanding of the structure and coherence of the creative performance. Semantic nodes can be represented or embodied as one or more data elements.

[0066] The simplified semantic network (using either topic modeling, or k- means clustering) can be used in combination with Graph Theory methods to extract additional features of creative performance, such as the number of semantic nodes used (clustering coefficient), the density of the space (average shortest path), and the modularity of the network.

[0067] Shape Uniqueness

[0068] Shape Uniqueness is a meta-property of Shape of Thought. It is the quantitative distinctness of a shape from other shapes (usually, but not always, the distinctness of the shape generated by one individual from the shapes generated by a group other individuals). Two ways in which shape uniqueness can be calculated are: (1) combining multiple features of a shape into a single model to compare that combination of features across multiple shapes, and (2) using Al based visual algorithms to identify shape uniqueness holistic visual shape uniqueness. Shape uniqueness can be represented or embodied as one or more data elements.

[0069] Shape of Thought Tasks

[0070] To create its SoT map, the Shape of Thought system can beoperable to execute computer executable instructions that use a battery of standard tasks to assess creativity and semantic generation. The tasks can be rated by human raters and by the Shape-of-Thought analysis. Any task that can include semantic generation can be used, however, the SoT system can perform the following tasks: a. The Alternative Uses Task (AUT). This task measured divergent thinking by asking participants to generate alternative uses for four items: fork, lamp, marble, and stick. The items varied in their semantic constraint, with fork and lamp being low-constraint and marble and stick being high-constraint. The items were presented randomly, and participants had two minutes per item. The Shape- of-Thought analysis calculated the Volume, FA, Idea Spacing, and Linearity for each item based on the participant’s responses. The items and analysis can be represented or embodied as one or more data elements. i. Evaluation of self-generated ideas. In the self-evaluation of ideas task, participants are presented with the alternative uses they themselves have produced during the AUT. The participants are asked to rate the ideas they produce for each object for their originality / creati vity / usefulness / appropriateness either by ranking them, or by assigning them with a score (1-5). The self-evaluation results can be represented or embodied as one or more data elements. ii. Evaluation of other-generated ideas. In the other-evaluation of ideas, participants go through a similar rating procedure, however in this case they rate ideas from an AUT produced by other people. Results of evaluation of other-generated ideas can be represented or embodied as one or more data elements. b. Forward Flow (FF). This task measured semantic fluency and flexibility by asking participants to generate a word that wasremotely associated with a given word and then to continue producing a chain (10-20 words) of word-pair associations. This task was scored on the ability of the participant to produce distant associations; a semantic space (i.e. Ada2, BERT) was used to translate word-pairs into semantic distance. This task is scored on Fluency (how many items produced), Flexibility (how many categories), and Originality (statistically infrequent). This task and its scores (e.g., fluency, originality) can be represented or embodied as one or more data elements. c. Creative Writing task. This task measured creative writing skills by asking participants to write a short story (around four sentences long) that included the following three words: stamp, send, and letter. The task duration was five minutes. Participants were instructed to use their imagination and to be creative when writing the story. The story and story elements can be represented or embodied as one or more data elements. d. Open-ended divergent thinking task (OEDT). This task measured open-ended divergent thinking by asking participants to imagine ways that transportation could be improved in the future to solve a problem or problems that exist now. The subjects were instructed to think creatively and to write down their entire thought process, without leaving out any association, no matter how strange it might seem. They were informed that the conductors of the task were not interested in the answer to the problem as much as to their thought process. The task duration was five minutes. Data associated with OEDT can be represented or embodied as one or more data elements. e. Fluency. This task measured verbal fluency by asking participants to write down as many instances that related to a certain prompt in two minutes. The Shape of Thought system can use, for example, two prompts: words that start with the letter P, and things that are bigger than your hand. Fluency can be represented or embodiedas one or more data elements.

[0071] In accordance with example embodiments of the present application, the Shape of Thought system includes dimension reduction to two or three dimensions, which are visible shapes that facilitate human interpretability, instead of operating within the high-dimensional semantic spaces of large language models, meaning that the only human-interpretable outcomes are numbers. Put another way, the Shape of Thought system can allow for visualizing and interacting with trajectories through semantic space, including responsiveness and interaction in real time (e.g. , to see how the volume of shape-of-thought changes as a person generates new ideas). This can be applied to training and gamification.

[0072] The Shape of Thought system can evaluate trajectory through semantic space as a trait of an individual, which can characterize individual differences in the thought processes of different people. The stability of Shape-of-Thought features (i.e. , as a consistent trait of a person’s way of thinking) has been empirically demonstrated, and links individual differences in Shape of Thought to individual differences in achievement, personality, and creative ability.

[0073] Further, the Shape of Thought system can use multiple features of thought trajectories in combination. Combinations of shape features can be used to predict outcomes related to personality, creative ability, and future achievement (e.g., using college admissions essays to predict the future college GPA of applicants).

[0074] In accordance with exemplary embodiments of the present application, specifically in the context of assessing creative thinking, the Shape of Thought system can focus on the spatial representation of creative thinking through geometric features derived from semantic spaces. This can offer a more nuanced understanding of the creative process. Unlike existing technologies that often rely solely on semantic analysis or quantitative metrics, the Shape of Thought system can integrate both semantic and geometric features. This dual approach can provide for a richer and more comprehensive evaluation of creativity, capturing both thecontent and structure of creative ideation.

[0075] In accordance with exemplary embodiments of the present application, the Shape of Thought system can also leverage machine learning algorithms to analyze individual differences in the geometric features of creative production. This dynamic integration enhances the predictive capabilities of the framework, offering a data-driven and personalized assessment of creative potential.

[0076] In accordance with exemplary embodiments of the present application, the Shape of Thought system can use multiple features of thought trajectories in combination. Using combinations of shape features facilitates prediction of outcomes such as personality, creative ability, and future achievement (e.g., using college admissions essays to predict the future college GPA of applicants).

[0077] In accordance with exemplary embodiments of the present application, the SoT system can allow a person to visualize the features of their thought process, and then interactively and dynamically allows them to develop their desired features. The SoT system can receive the person’s verbal input, analyze it, and then display that person’s shape of thought features, which can be presented on a visual interface. The person, upon seeing the representation of his or her thought, can now, taking the visual representation as feedback, vary or adjust their thinking, and provide a subsequent verbal input. The subsequent verbal input can be analyzed and the SoT system can provide an updated SoT visualization that reflects the subsequent verbal input. Thus, the person can respond to an output by providing a subsequent verbal input, the generated visualization of which can then be displayed to the person in real time (or near real time), which may be after a short period, or perceived by the person to be immediate. In this responsive, interactive, and dynamic manner, the person is able to train their mind to think in a way that can generate patterns indicative of, for example, an art, a skill, problem solving ability, and the like.

[0078] Referring now to FIG. 9, a computer 900 (computing device, machine, desktop, laptop, tablet, etc.) can be used to implement exemplaryembodiments of the Shape of Thought, and can include electronics-related components that enable Shape of Thought functionality and operations, including the methods described and claimed below. In each of the example operations described above, steps, blocks, features, or aspects can be substituted or combined with other steps, blocks, features, or aspects, unless context warrants that such combinations or substitutions are not possible. Further, if a step, block, feature, or aspect is not described with respect to example operations, this does not mean that the feature, step, or aspect is incompatible or impossible with respect to those example operations. As such, the example operations of the present application described above are not necessarily limited to the steps, features, or aspects that are described with respect to those example operations.

[0079] Computer 900 can comprise a microprocessor 904 (of referred to as a processing unit, central processing unit, processor), which can be one or more microprocessor chips (including AICS). The microprocessor 904 (which can also be one or more microprocessors) can execute machine executable instructions (e.g., computer program product, computer- readable instructions, software, software programs, software applications, software modules, etc.) to facilitate performance of the operations (e.g., functions, methods, processes, and logic flows) described herein. The microprocessor 904 can be any of various commercially available processors capable of performing the functions described herein, and can include brands sold by Intel, AMD, etc.

[0080] The computer 900 can also comprise a computer readable media, such as system memory 906, which can include one or more memory components, such as read only memory ROM 910, a non-volatile memory, and random-access memory RAM 912. The ROM 910 can store, for example, a basic input / output system (BIOS), an operating system 930, applications 932 of the computer 900, and other computer modules 934, and generated data 936. The RAM 912, which is often called “working memory,” can also include a high-speed RAM such as static RAM for caching data. Other tangible and / or non-transitory media can be used tostore desired information. In this regard, the terms “tangible” or“non- transitory” herein as applied to storage, memory or computer-readable media, are to be understood to exclude only propagating transitory signals per se as modifiers and do not relinquish rights to all standard storage, memory or computer-readable media that are not only propagating transitory signals per se.

[0081] Still referring to FIG. 9, a system bus 908 can connect system components including, but not limited to, the system memory 906 to the microprocessor 904. The system bus 908 can be any of several types of bus structure that can further interconnect to a memory bus (with or without a memory controller), a peripheral bus, and a local bus using any of a variety of commercially available bus architectures.

[0082] The computer 900 can also comprise an internal hard disk drive (HDD) 914 (e.g., EIDE, SATA), which can be connected via a storage interface 924 to the system bus 908. The computer 900 can also comprise an interface (e.g., USB interface) with any number of external storage devices (e.g., a memory stick or flash drive reader can allow a flash drive to be connected to the computer, to facilitate printing of documents stored on the flash drive). Additionally, a solid-state drive (SSD) could be used in addition to, or in place of, an HDD 914.

[0083] A user can enter commands and information into the computer 900 using a user input interface, labeled in FIG. 9 as user input interface 910. The user interface 910 can comprise a keypad 938 and a touch screen 940, and the user input interface 910 can be coupled to the system bus 908 via an input device interface 944.

[0084] The computer 900 can operate in a networked environment using logical connections via wired and / or wireless communications via a communications network 956 to one or more remote computers, such as a remote computer 950. The remote computer 950 can be a server with access to networked storage (e.g., cloud storage devices), a desktop computer, laptop, smartphone, tablet, or the like. As such, in some example embodiments in accordance with the present application, data stored in, forexample ROM 910 can also be stored remotely and accessible on-line. The communications network can be a local area network (LAN, which wired or wireless network, such as a Wi-Fi network. The communications network can also comprise larger networks (e.g., a wide area network (WAN)) that facilitate enterprise-wide computer networks, such as intranets, all of which can connect to a global communications network, e.g., the Internet. A wireless communication network interface or network adapter 958 can facilitate wired or wireless communication to the communications network 956.

[0085] As used herein, the singular terms "a," "an," and "the" may include plural referents unless the context clearly dictates otherwise. Reference to an object in the singular is not intended to mean "one and only one" unless explicitly so stated, but rather "one or more."

[0086] As used herein, the term "set" refers to a collection of one or more objects. Thus, for example, a set of objects can include a single object or multiple objects.

[0087] As used herein, the terms "substantially" and "about" are used to describe and account for small variations. When used in conjunction with an event or circumstance, the terms can refer to instances in which the event or circumstance occurs precisely as well as instances in which the event or circumstance occurs to a close approximation. When used in conjunction with a numerical value, the terms can refer to a range of variation of less than or equal to ±10% of that numerical value, such as less than or equal to ±5%, less than or equal to ±4%, less than or equal to ±3%, less than or equal to ±2%, less than or equal to ±1 %, less than or equal to ±0.5%, less than or equal to ±0.1 %, or less than or equal to ±0.05%. For example, "substantially" aligned can refer to a range of angular variation of less than or equal to ±10°, such as less than or equal to ±5°, less than or equal to ±4°, less than or equal to ±3°, less than or equal to ±2°, less than or equal to ±1 °, less than or equal to ±0.5°, less than or equal to ±0.1 °, or less than or equal to ±0.05°.

[0088] Additionally, amounts, ratios, and other numerical values maysometimes be presented herein in a range format. It is to be understood that such range format is used for convenience and brevity and should be understood flexibly to include numerical values explicitly specified as limits of a range, but also to include all individual numerical values or sub-ranges encompassed within that range as if each numerical value and sub-range is explicitly specified. For example, a ratio in the range of about 1 to about 200 should be understood to include the explicitly recited limits of about 1 and about 200, but also to include individual ratios such as about 2, about 3, and about 4, and sub-ranges such as about 10 to about 50, about 20 to about 100, and so forth.

[0089] Although the description herein contains many details, these should not be construed as limiting the scope of the disclosure but as merely providing illustrations of some of the presently preferred embodiments. Therefore, it will be appreciated that the scope of the disclosure fully encompasses other embodiments which may become obvious to those skilled in the art.

[0090] All structural and functional equivalents to the elements of the disclosed embodiments that are known to those of ordinary skill in the art are expressly incorporated herein by reference and are intended to be encompassed by the present claims. Furthermore, no element, component, or method step in the present disclosure is intended to be dedicated to the public regardless of whether the element, component, or method step is explicitly recited in the claims. No claim element herein is to be construed as a "means plus function" element unless the element is expressly recited using the phrase "means for". No claim element herein is to be construed as a "step plus function" element unless the element is expressly recited using the phrase "step for".

Claims

CLAIMSWhat is claimed is:1 . A data processing and display method comprising: receiving, by a device comprising a processor and memory, a plurality of first verbal data inputs, wherein: each of the plurality of first verbal data inputs are reflective of the subject’s thought, and the first verbal data inputs comprise words or sentences; based on the plurality of first verbal data inputs, determining, using an analysis module executed by the device, a plurality of data elements comprising one or more of: an idea spacing data element that relates to a semantic diversity measurement and a novelty of ideas measurement, a volume data element that relates to a semantic extent of richness of creative performance measurement, a fractional anisotropy data element that relates to a semantic directionality and coherence of creative performance measurement, a linearity data element that relates to a semantic progression and continuity of creative performance measurement, a semantic network nodes and edges data element related to words or sentences derived from the plurality of first data inputs, an optimal path data element indicative of a density related to a semantic space, a topic modeling data element comprising an arrangement of words or sentences around topics or themes based on their semantic similarity, and a shape uniqueness data element reflective of a quantitative distinctness of a shape from other shapes;mapping, by the device, a trajectory based on the plurality of data elements, wherein the trajectory comprises a geometric feature displayable on a display device; and receiving, by the device, a plurality of second verbal data inputs related to a response to an evaluation of the trajectory by the subject, wherein the plurality of second verbal data inputs is representative of a second set of words or sentences.

2. The method of claim 1 , further comprising: based on the plurality of second verbal data inputs, determining, using the analysis module, another plurality of data elements comprising one or more of: another idea spacing data element, another volume data element, another fractional anisotropy data element, another linearity data element, another semantic network nodes and edges data element, another optimal path data element, another topic modeling data element, and another shape uniqueness data element; mapping, by the device, another trajectory based on the another plurality of data elements, wherein the another trajectory comprises another geometric feature displayable on the display device; and facilitating generating, by the device, the another trajectory on the display device.

3. The method of claim 1 , wherein the idea spacing data element is derived from a calculation that is an average cosine distance between an embedding of the plurality of first verbal data inputs in a two or three- dimensional space. re4. The method of claim 1 , wherein the idea spacing data element of the plurality of data elements is indicative of a distant and diverse idea.

5. The method of claim 1 , wherein the idea spacing data element of the plurality of data elements is indicative of a similar or conventional idea.

6. The method of claim 1 , wherein the volume data element is derived from a calculation that is the minimum volume of a multi-dimensional ellipsoid that encompasses an embedding of the plurality of first verbal data inputs in the semantic space.

7. The method of claim 1 , wherein the volume data element of the plurality of data elements is indicative of a creative performance measurement that is semantically large and rich.

8. The method of claim 1 , wherein the volume data element of the plurality of data elements is indicative of the creative performance measurement that is semantically limited and narrow.

9. The method of claim 1 , wherein the fractional anisotropy data element is derived from a calculation as a ratio of a largest eigenvalue to the sum of all eigenvalues of a covariance matrix of embeddings of the plurality of first data inputs in the three-dimensional space.

10. The method of claim 1 , wherein the fractional anisotropy data element of the plurality of data elements comprises a value of 0 indicative of an isotropic semantic space that is representative of a creative performance measurement with theme and direction.

11. The method of claim 1 , wherein the fractional anisotropy data element of the plurality of data elements comprises a value of 1 indicative of ananisotropic semantic space that is representative of the creative performance measurement that is exploratory and diverse.

12. The method of claim 1 , wherein the linearity data element is derived from a calculation of a correlation between an order of the plurality of first data elements and angles between responses, derived from embeddings in a two or three-dimensional space.

13. The method of claim 1 , wherein the linearity data element of the plurality of data elements is indicative of a creative performance that follows a linear semantic trajectory that is representative of the creative performance being logical and consistent.

14. The method of claim 1 , wherein the linearity data element of the plurality of data elements is indicative of a creative performance that deviates from the linear semantic trajectory and is representative of the creative performance being surprising and unexpected.

15. The method of claim 1 , wherein the semantic network nodes and edges data element utilizes graph theory methods to extract additional features.

16. The method of claim 1 , wherein the optimal path data element is calculated as an average of shortest paths between all pairs of nodes.

17. The method of claim 1 , wherein the topic modeling data element relates to additional features such as number of nodes / topics, themes, and semantic nodes.

18. The method of claim 1 , wherein the shape uniqueness data element can be calculated by combining multiple features of the shape into a single model to compare that combination of features across multiple shapes.

19. The method of claim 1 , wherein the shape uniqueness data element can be calculated using Al based visual algorithms to identify shape uniqueness.

20. A computer system comprising a processor and memory, a processor; and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, comprising the method of claims 1-19.

21. A non-transitory computer-readable medium storing instructions that, when executed by a device comprising a microprocessor, causes the device to perform the method of claims 1-19.

22. A system comprising: a processor; a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations comprising: receiving, by the system, a plurality of first verbal data inputs, wherein: each of the plurality of first verbal data inputs are reflective of the subject’s thought, and the first verbal data inputs comprise words or sentences; based on the plurality of first verbal data inputs, determining, using an analysis module executed by the system, a plurality of data elements comprising one or more of: an idea spacing data element that relates to a semantic diversity measurement and a novelty of ideas measurement, a volume data element that relates to a semantic extent of richness of creative performance measurement, a fractional anisotropy data element that relates to a semantic directionality andcoherence of creative performance measurement, a linearity data element that relates to a semantic progression and continuity of creative performance measurement, a semantic network nodes and edges data element related to words or sentences derived from the plurality of first data inputs, an optimal path data element indicative of a density related to a semantic space, a topic modeling data element comprising an arrangement of words or sentences around topics or themes based on their semantic similarity, and a shape uniqueness data element reflective of a quantitative distinctness of a shape from other shapes; mapping, by the system, a trajectory based on the plurality of data elements, wherein the trajectory comprises a geometric feature displayable on a display device; and receiving, by the system, a plurality of second verbal data inputs related to a response to an evaluation of the trajectory by the subject, wherein the plurality of second verbal data inputs is representative of a second set of words or sentences.

23. The system claim 22, wherein the operations further comprise: based on the plurality of second verbal data inputs, determining, using the analysis module, another plurality of data elements comprising one or more of: another idea spacing data element,another volume data element, another fractional anisotropy data element, another linearity data element, another semantic network nodes and edges data element, another optimal path data element, another topic modeling data element, and another shape uniqueness data element; mapping, by the system, another trajectory based on the another plurality of data elements, wherein the another trajectory comprises another geometric feature displayable on the display device; and facilitating generating, by the system, the another trajectory on the display device.

Images