Informations

Structure of Corpus

Click in the model area to enlarge it.

Structure of Data Implementation into Databases

Click in the model area to enlarge it.

Fictional Topography Models – Character Wandering Network Models - Place Density - Character Networks

The models use vectorised maps that are conceived in Adobe Illustrator. The purpose of these maps is to provide a basic topographic fundanet into which layers of fictional topographies, network models and site density models are fed. Caracter Wandering Network Models of character journeys show the connections between the places where a character or characters move. You can click on Network Model Type Visualization to see the network model types. Place desnity map shows each location according to its frequency load in a given text. The size of the circle is calculated as 2πr, where r is the relative frequency of the toponym. Character Networks illustrate the relationships and interactions between different characters within the narrative space.

Number of Residents and Houses in Prague and Surrounding Districts in 1869-1950

Prague's topography hardly changed during the 19th century, which cannot be said about the population and especially the new houses that grew up in the districts that were built outside the original Prague walls during the 19th century. At the turn of the century, the so-called Prague Redevelopment took place and in 1922 the so-called Greater Prague was declared.

Frequency of Toponyms and Locations – Prague and Non-Prague Locations – Sacred Spaces – GIS

The toponym frequency graphs are obtained by filtering toponyms from a special TXT file that is manually extracted from the texts. Automatic detection is also used using the so-called named entity recognition method (for Czech names, see Czech Named Entity Corpus 1.0). The frequencies of different ways of naming a toponym are grouped under one name; for example, in Arbes's Saint Xavier we find the following ways of naming the St. Nicholas Cathedral in the Lesser Town: Mikuláš, Mikuláš Cathedral, Malostranská Cathedral. By sites we generally mean Prague and non-Prague sites. Here, we distinguish whether the action takes place in Prague (in this case, regardless of the specific Prague topographical designation) or outside Prague. This criterion varies, of course, depending on the urban development of the city, especially since the end of the 19th century. In an interpretation that takes this aspect into account, it is therefore important to consider individual sites in relation to the historical changes in Prague's topography. For example, Smíchov in Arbes's novels represents a space beyond the borders of Prague. Sacred places are categorized as: temple, ancient temple, cathedral, monastery, church, parts of church, church interior, rectory and cemetery. These categories are selected according to entries from the Thesaurus of the Czech Language. GIS maps associate fictional topographic locations with real places according to coordinates (longitude, latitude). Places that still exist are distinguished from places that have disappeared.

Text Annotation

All texts in the corpus are annotated not only morphologically using the Morphodita tool, but also with special tags to mark individual text segments and locations. An overview of these tags is provided in the table below.

Statistics

This section contains information regarding the text segments that were analyzed. They are listed on the y axis of the Sentence Lengths chart, which shows the average sentence lengths of single-character segments as a function of the specific text, as well as across the entire author coprus (lower area of the chart). The second graph contains boxplots (box plots) that show the dispersion of text by sentence lengths. A boxplot is a representation of the values between the first (Q₂₅) and third quartiles (Q₇₅), i.e., the variance between 25% - 75% of the values. Inside the boxplot is a meridian (Q₅₀) showing the mean value(s) in the set of numerical sequences. Outliers are also part of the plots.

Example:

1, 2, 5 ,8, 15, 16, 16, 17, 18, 20, 25, 30
An ordered numeric series of measured/acquired values (in our case, this would be the individual sentence lengths in the text) has a size of 12 items. Q₂₅ = 3, Q₅₀ = 6, Q₇₅ = 9. The boxplot will include the area between the intervals Q₂₅ and Q₇₅, which means the numbers: 5 , 8, 15, 16, 17, 18, where Q₅₀ = 16. In an ordered set of values, the boxplot is the expression of those values that lie in this interval. The meridian is the mean, not the arithmetic mean, which in this case is 14.4. The resulting boxplot shows a higher concentration of values in the 3rd quartile, i.e. in the range between 50% - 75%:

Entropy expresses the so-called degree of uncertainty of the system or the degree of its disorder. Entropy is directly related to the probability of occurrence of elements in the system. in other words, the higher the probability of the elements of the system, the lower the degree of its uncertainty and vice versa. In information theory, the following formula so called Sahnnon entropy is used to calculate entropy:

$$H = - \sum_{i=1}^{n} p_i \log_2(p_i)$$

The probability of the element p(i) is given by: $$P(i) = \frac{\text{Af(i)}}{\text{N}}$$

Af(i) is the absolute frequency of a given element (e.g., word) and N is the size of the text (e.g., calculated per word count). The relationship between probability, entropy, frequency and semantics of a linguistic feature can be expressed as follows:

$$Z^{(-\inf, +\text{sem})}_{freq+} = \text{low entropy}$$ $$Z^{(+\inf, -\text{sem})}_{freq-} = \text{high entropy}$$

inf = information saturationt
sem = semantics
freq = freqency

The more frequent the feature, the lower its information saturation and the lower the entropy value and vice versa. If the entropy is 0, it means that the system is completely ordered, i.e. 100% predictable. These three short versions of random texts may serve as an example (texts was generated by Chat GPT 4o):

To calculate lexical diversity and display it in a graph with linear regression, at least two paired values (x, y) are required (This is the reason why such graphs are missing for some authors, as their corpus currently contains only a single work.), where x represents unique words and y represents all words. The calculation uses the formula:

$$y = mx + b$$

where m is the slope (gradient of the line) and b is the offset (y-intercept of the line). The slope is calculated as:

$$m = (y2 - y1) / (x2 - x1)$$

For the values x1, y1 = 5, 23 and x2, y2 = 3, 58, the slope is -17.5. The offset is calculated as:

$$b = y1 - m * x1$$

With the given values, the offset is equal to 110.5.

The readability of the text is determined by:

Sentiment Analysis – Cluster Sentiment – Word Clouds – Cluster Sentiment Graph Line – MDS

Sentiment analysis is carried out on the basis of the list of lemmas made available by Kateřina Veselovská and Ondřej Bojar within the project SubLex 1.0. In the course of the Czech prose corpus project, we are currently trying to work with the lexicon excerpted from the Thesaurus of the Czech Language, specifically with entries containing lexemes for expressing emotions. Their provisional listing is available in Open Data. Word cluster models are the set of the most frequent autosemantics in a given work. The size of a word in the model corresponds to its frequency load. In particular, synsemantics or other words are removed during the analysis. Their list is in the Open Data section. Clustering models are used to track the main motifs. The Clusters of Sentiment section allows you to search for thematic clusters of emotions in each prose of the author corpus. Clusters are defined according to the "Thesaurus of the Czech Language" by Aleš Klégr et al.

The structure of JSON file containing dictionary for thematic sentiment analysis:

{"NAME OF EMOTIONAL" : [emotional lemmas]} Cluster Sentiment Graph Line display the comparable frequency of the individual clusters of emotions in the texts. The graph is based on the frequency of the lemmas in the text. The MDS (Multidimensional Scaling) model is a two-dimensional representation of the clusters of emotions in the text. The graph shows the distances between texts with regard to their sentiment cluster distribution. The closer the texts are to each other, the more similar the frequency distribution of their sentiment clusters is.

The cumulative frequency shows the occurrence of lemmas with positive and negative emotionality in the text. The curve only counts these lemmas, i.e. its length (x-axis) depends on the size of the dictionary containing such lemmas, i.e. 5685 lemmas, not on the length of the text. The y-axis shows the cumulative frequency of these lemmas in the text. The curve thus shows how many of these lemmas are found in the text. The steeper the curve, the more frequent the occurrence of these lemmas in the text. If the curve is flat, it means that these lemmas do not occur in the text at all or only very rarely. However, this is not the case in the corpus.

Search Word Types by Simple Tag

The search for word types is performed in a database of morphologically tagged texts using the MorhoDita tool developed by Milan and Jana Straka. After entering a collapsed tag (the first position in a 15-position tag) for a particular word species, the corresponding lemmas are searched, sorted in descending order of frequency. The lemmatized and morphologically tagged text files are accessible in Open Data.

CQL

CQL is a corpus query language. For more informations click here .

Search examples:
[lemma="something"] (searches for lemmas containing something)
[tag="something"] (searches for tags containing something)
[word="something"] (searches for words containing something)

[word="*ti"] (searches for words ending in -ti)
[lemma="*ti"] (searches for lemmas ending in -ti)

[word="something"] OR [lemma="*ti"] OR [tag="something"]