**4. New approaches to the assessment of domain-specific creativity**

Computer-based analyses allow for objective methods to make an assessment of the semantic-linguistic quality of narratives at the text level. In this regard, researchers have begun to explore the benefits of automated scoring approaches using computer-based computational tools [33, 61–67]. One of these approaches is Latent Semantic Analysis (LSA).

Being one of the automatic scoring approaches LSA, is a technique to extract and represent sentences with mathematical or statistical calculation. LSA express the ideas about that the similarities and differences in the meaning of words can be influenced by the similarities and differences in the overall context in which the word is there or not [68].

On the other hand, LSA is a language learning model in which meanings of words are extracted from statistical analyses of large chunks of text. LSA determines whether the words are related or not through analysis of the relationship between words based on which words are frequently used together and which words are rarely together [57, 69].

In addition, LSA is also used to measure the consistency of texts [70]. The aim of LSA is to create a structure that shows the level of similarity between words [69, 71]. With LSA, hidden connections in textual data are revealed. Unstructured data must be converted to structured format so that LSA can be used. Thus, LSA can be applied to any document stack whose syntactic and grammatical structure is cleaned [72]. Also, Heinen and Johnson [63] found that LSA-based measures of semantic distance relate to measures of novelty and appropriateness, measures of creative output [34, 67].

### **4.1 Using semantic distance in assessing creativity**

The semantic distance plays a role in various models about creative process research in field. So, Latent Semantic Analysis is a popular computational method to represent semantic distance in creativity research is through [65, 66, 69]. The role of semantic distance in creativity is rooted in the associative theory of creativity [10, 65, 73]. Therefore, the further a new concept moves away from a concept in a semantic space, the newer or more creative it will be. This is consistent with the "Associative Theory of Creativity," which states that creativity involves the ability to connect relatively weak or distant concepts to each other and to combine them with

*Perspective Chapter: New Approaches to the Assessment of Domain-Specific Creativity DOI: http://dx.doi.org/10.5772/intechopen.102311*

new and useful objects. That is, the greater distance there is between the concepts produced, the newer combinations are produced and the more creative the produced thing is [10, 65, 69].

Creativity involves the ability to associate relatively weak or distant semantic components and to combine them into new and useful objects [74]. Researchers focusing on creativity have long struggled with how to measure creativity [56]. Examining studies showing that creativity can be achieved through computational processes, Boden [19] argued that it may be possible to create a program that can create works of art or symphonies [36]. The literature contains studies suggesting that computer-based and artificial intelligence supported creative products can be developed [75–77] on the one hand and studies on whether creativity can be assessed automatically [57, 58, 78] on the other hand. However, it is stated that there are limited studies on measuring creativity based on semantic distance [74]. For example, the creativity literature does not include much research on estimating the originality of an idea in a written work [44]. It is seen that there are just a couple of studies, and they deal with behavior [79] and words or expressions as a creative way of using language [80]. In addition, other studies in the literature show that writing quality is associated with stronger reading skills [81–83], broader vocabulary [81, 84], grade level [85], more flexible thinking skill [86], and level of knowledge on the topic to be written about [87].

In the study conducted by Runco, Turkman, Acar, & Nural [88], one of the more recent studies in the literature, the relationship between idea frequency and creativity is investigated. Another recent study [44] has incorporated new measurement techniques (keyword study) into the assessment process of linguistic creativity. With the computer-based keyword method used, yet another study [88] demonstrated how creativity revealed itself in written works (products) and to what extent this method reduced the time and effort cost in scoring multiple thinking tasks. Furthermore, the tests administered to the participants were assessed by domain experts (subjective) along with computer-based (objective and automatic) assessment, the new method. The obtained assessment results were promising for the objective assessment in hand.

Researchers in the literature [89, 90] found that divergent thinking (DT) tasks can be properly scored using LSA [72, 91]. The semantic word categorization performance of LSA [92] is reported to be satisfactory and comparable to human performance [65]. Additionally, LSA is preferred as it is objective, not based on human judgment, which allows obtaining reliable results among users, measurable and numerically applicable, grounded on a theoretical background, and justified [93] and it offers a strong way for quantitative analysis [73]. It can be stated that these findings show the usefulness and effectiveness of semantic distance measures to measure domain-specific creativity.

### **5. Method**

This chapter is a descriptive study. The method of research which concerns itself with the present phenomena in terms of conditions, practices beliefs, processes, relationships or trends invariably is termed as "descriptive survey study" [94]. In addition, descriptive research is devoted to the gathering of information about prevailing conditions or situations for the purpose of description and interpretation. This type of research method is not simply amassing and tabulating facts but includes proper analyses, interpretation, comparisons, identification of trends and relationships. It is

concerned with the present and attempts to determine the status of the phenomenon under investigation. The survey research employs applications of scientific method by critically analyzing and examining the source materials, by analyzing and interpreting data, and by arriving at generalization and prediction [95].

According to Pandey and Pandey [95], Types of Descriptive Method: Descriptive method is divided into four parts. You may view the schematic below (**Figure 4**).
