platform for the analysis of written texts in terms of .../sci/pdfs/zs050dd.pdf · [16]. past...

Platform for the analysis of written texts in terms of conceptual graphs: study

case, Cybernetics of Cybernetics course

José Bermeo

Germán Bravo-Córdoba

Sebastián Contreras

Roberto Zarama

Universidad de Los Andes

Cra. 1 Este 19 A- 40

Bogotá - COLOMBIA

{j-bermeo, gbravo, seb-cont, rzarama}@uniandes.edu.co

ABSTRACT

The objective of this work is to observe and analyze the

conceptual-relation structures in texts written in Spanish, from

the perspective of second-order cybernetics. Texts are shaped

by the syntactic structure of the sentences they contain.

Conceptual-relationships emerge by transforming the text, using

a text analysis platform named PAST. The transformation is

guided by a set of established rules, permitting to observe a

system. PAST can be defined as a powerful operator to

transform texts written by the students in a Master’s Program

class. The platform is introduced as a tool enabling a graph

representation of relationships among the concepts in the text.

The internal procedure of the platform and the graph

construction concerns the linguistic analysis. Although the

resulting graphic representation does not follow a formal

definition, it reflects an iterative construction to define a

topology to create a semantic network. This network only

makes sense to the author of the text. In other words, the author

(observer), based on the cybernetics of the observing systems,

uses a tool that enables him to make a second-order observation

of a document that he/she has linearly written. The main

epistemological concepts mentioned throughout the

development of the proposal are discourse and distinction. The

contribution of this work is the application of a methodology

that conducts a recursive observation of both the direct and

indirect relationships established by the words the author puts

together in his/her written discourse. In the discussion of results

the next phases of this work are presented.

Keywords: Cybernetics, Observer, speech, distinction,

semantics, conceptual graph.

1. INTRODUCTION

The objective of a Cybernetics of Cybernetics course (CC) is

the questioning, reflection, and construction about the

observer’s observing: The interest is that students in the course

become able to see the process they follow when they do the

observations they do. Such actions are carried out from the

perspective of language, questioning the observer who observes

the systems, stating his/her own purpose. Regarding the

language, the actions of the Observer on first-order cybernetics,

ending in second-order cybernetics, are referred to as observer

of the observer [6]. The main question that defines the course’s

core: Who is observing me as an observer? To answer this

question, a methodology was designed, involving games and the

construction of a space allowing the interaction among the

members of a system (in this case, the CC course), and the

theory of graphs [7]. Games are the main pillar of the proposed

methodology and it is also the tool used to implement the

methodology. Games are understood as the action defining a

space where a “world” emerges, bounded by its interpretative

context and its own rules.

The CC course defines learning as the acquisition of different

concepts and their interconnections, which are corporally

absorbed. In the case of CC, the learning and absorbing

processes occur as a result of recurrent reading, writing essays,

and playing games. The subject of readings is the Observer,

his/her emotions and perspectives. In order to analyze the

construction of the game space for action and the relationships

generated by the methodology, the concept of social networks is

used: the design and organization of the players (CC course

students) is defined by a regular polyhedron.

During the qualitative revision of the students’ essays, the need

to create a text-analysis platform supporting this revision

process, while maintaining the author’s (student’s) style and

relationship among concepts intact. A first approach used an

automated tool to observe and measure structural changes,

based on the words and their frequency in the students’ essays

[16]. PAST presents then the students’ texts in word-

relationships through a graphic representation of the text. The

platform transforms the original text into a scheme discovering

the relationships among its words. With the word relationships,

a graph is depicted as a response file. Thus, we obtain a

structure reflecting the relations among the words (nodes),

based on linguistic dependencies and in the rules to structure the

links (edges). The elements and relations in the graph are

interpreted as the distinctive features the author uses to write a

text in every single opportunity. Besides of a a different

observation of the essay, this analysis also favors the course

objective by encouraging the linking of different concepts

learned during lectures.

The focus of this document is the text-analysis platform

(PAST). To accomplish this objective, some formalization

SYSTEMICS, CYBERNETICS AND INFORMATICS VOLUME 7 - NUMBER 1 - YEAR 2009 41ISSN: 1690-4524

principles are required: 1. the author (observer) do re-construct

his/her written discourse in three opportunities; 2. The platform

offers the possibility to build networks of interconnected words.

The platform is based on a linguistic context, and a set of rules

to identify and relate the networks of words; 3. Rules are

defined by the user (professor, analyst and student) and they

encourage the generation and extraction of personalized

structures. The grouping of such structures constitutes the words

network; 4. The suggested transformation to represent the

conceptual relations in texts implies the use of graphs as the

main data structures in the platform; 5. The main concepts

identified are interpreted as the distinctions the author uses in

the language of his/her text.

Once formalized these principles, the document is divided in

five sections: 1. the epistemological considerations framing the

work; 2. the description of PAST as the text analysis platform

for the CC course (PAST); 3. the application of PAST and the

issued results, including the rules for the analysis, application of

the graph and visualization; 4. Discussion of the results; and 5.

conclusions.

2. EPISTEMOLOGICAL CONSIDERATIONS

The written text in association with the discourse

According to Foucault, in spoken discourse "The production of

discourse is not alienated from the social system in which it is

produced” [12]. The CC course defines a social mechanism to

regulate the production of discourse and set the boundaries of

the space where it is produced. Von Bertalanffy states that the

relationship between the language and the vision of the world is

not a one-way relationship but a reciprocal one, according to his

words "the language structure seems to determine those

characteristics of reality to be abstracted and also how thinking

categories will be shaped. Thus, the perception of the world

determines and shapes the language" [8]. In an adaptation of

Bertanlanffy’s words, for the observer’s understanding, the

discourse (understood as language) expresses the shape of the

particular system, subject of the discourse. However, when our

discourse is written in the western way: “(…) our language is

presented as a causal vision of linear nature” [3]. In other

words, it seems that: “we accept the language as an instrument

that determines our thoughts and experiences” [2], but an

individual’s discourse inevitably reflects the interpretative

context of the author as an observer, within the framework of

relations of the observer’s social structure. In theory, grammar

does not prescribe what we do [10], yet it describes what we say

as individuals.

The exercise proposed in the CC class and the use of PAST

transcends its implications within individual and collective

discourse, which arise as entities during the development of the

exercise. The use of the associated objects and actions implies

the combination and relationships of several words, each

constituting a unique referent (node). Then, those referents are

combined and related to others throughout arcs; therefore, the

resulting idea is a recurrent interaction process in which

language acts as the social system operator, independently of the

meaning of words. Thus, the meaning of propositions “are

determined by its use” [10]: That is, the meaning the observer

gives them when using them.

The concept of Distinction applied to the text

Spencer- Brown considers that the starting point of all fields of

knowledge requires inventing and making distinctions: "a

universe is born when it is separated or alienated from a space

and we can draw the limits anywhere we please" [4]. Distinction

constitutes the observing act that defines the frontier dividing a

space into two sub-spaces, two completely separated continents.

We associate the concept of distinction with the concept of

‘reference’ presented in Ogden and Richard’s semantic triangle,

which consists in the association among the referent (object),

the symbol (in this case the word linked to the object) and the

reference between the referent and the symbol. According to the

authors, a direct relationship between the referent and the

symbol cannot be made without the act of making a reference.

In consequence, the described process involves the act of

drawing a distinction in the language [Ogden and Richards cited

in 13].

In addition, the relationship of the words evidences the

individual’s cognitive structure in the language, wherein the act

of drawing a distinction in the cognitive level, corresponds to

the individual’s concept associations in different levels of

abstraction: “For example, the concept “tree” is connected to the

more abstract –more general- concept “vegetable”; that is, a tree

is a particular case of a vegetable or both. Likewise, the abstract

concept “animal” is connected to the more concrete –more

specific- “mammal,” and this one to the even more concrete –

more specific- “mouse” concept” [17]. Furthermore, a

representation of concepts linked by symbols (words of the text)

in a non linear way (conceptual graphs), constitutes a cognitive

scheme of such abstractions.

Conceptual graphs reflect the distinctions made by each student

(observer) each time the submission of an essay is required. In

the nodes (unary referent) and arcs topology, distinctions only

represent the language associations made by the observer.

Therefore, the operations of individual distinction are shaped

between networks of observers.

Components of grammar mastering

A number of theories try to explain the grammatical formation

of communication through language; in this document, we will

refer to “generative grammar”, as proposed by Noam

Chomsky. This author states that humankind is homogeneous in

terms of linguistic expression, like it is in terms of biology, and

that the linguistics of each spoken language is nothing but a

combination of basic structures of communication mechanisms.

The only genuinely linguistic dominion is the grammatical one,

which has three components: a syntactic component, a semantic

component, and a phonologic component [15].

Syntax is primordial and it generates the structures

(grammatical sentences). Grammatical sentences are made up of

individual words and their connections; such distinctions are

named constitutive grammar and grammatical dependence

respectively. Syntax is built by a basis and a transformational

process. The basis is the set of writing rules that generates

structure (a group of writing rules that define grammatical

relationships and a lexicon that defines the terms and adds

semantic and grammatical information). Transformations are

the rules that will turn deep structures into superficial structures.

The semantic component defines the meaning of those

structures, that is to say, the network of words with semantic

SYSTEMICS, CYBERNETICS AND INFORMATICS VOLUME 7 - NUMBER 1 - YEAR 200942 ISSN: 1690-4524

relationships, and the phonologic component makes those

structures perceptible for interpretation [1]. A hierarchic

representation of these grammatical components establishes a

dependence hierarchy as follows: syntactic - semantic and

phonological component. An additional component,

morphologic, has been proposed by L. Boomfield [9].

3. PRESENTATION AND TRANSFORMATION

OF THE TEXT ANALYSIS PLATFORM FOR

THE CC COURSE (PAST)

PAST is the platform designed to analyze the essays written in

the Cybernetics of Cybernetics class 2007. A fraction of the

reconstruction of the observer’s experience when observing the

CC course is expressed by writing essays. The starting text

corresponds to the essays written in the course. Essays are

reviewed and returned to students in three different times (t)1.

This action requires two operators. The first operator is iterative

and it occurs when: “applying the same operation on the same

subject once and again until it is operationally closed and

becomes the subject’s own behavior or value” [2]. The second

operator, the recurrent one: "[...] shows a system incorporated

within another [...], each one with a unique identity, despite the

incorporation” [18].

Given the course objectives as the questioning, reflection and

construction of the observer’s observer, after the professor’s

revision and comments of each of the three essays, the analysis

platform will enable the student (observer) to find non-lineal

relations in his written discourse. The platform supports and

facilitates the evaluation of essays by the professor, and also

helps to attain the course’s objective by letting students observe

the individual and collective construction of the observer’s

observer they are.

The process associated to the text transformation

This process builds a sequence of structures representing the

text content in different ways, each one evidencing some of its

characteristics. In this section, the transformations obtained in

each phase of the process are globally explained:

0. The first step is to eliminate styles, figures, charts, indices,

references, etc., leading to a plain text file of the document

(1).

1. The first transformation creates a representation of the text

maintaining the relationships among words. This

transformation only includes linguistic structures existing

in the text. By using of a syntactic analyzer called Freeling,

a set of linguistic dependence trees are formed, one for

each sentence of the text. Each node in the tree contains a

word, its associated motto, and its morphologic category,

represented by eagle labels; the arcs represent syntactic

dependencies, of the “adjective-noun”, “verb-noun” type.

2. The second transformation seeks the full representation of

the text and not of its separated sentences, taking into

account the relationships existing among the dependence

trees. The transformation of these relationships is based on

a set of rules of morphologic composition and

1 Course experience suggests a third iteration as the ideal number of

repetitions of an essay. Additional iterations may occur, but it would

cost, in terms of time, an occasional extra revision.

transformation (R1), which generate a graph of

morphologic structures when applied to each tree.

3. The rules for morphologic composition and transformation

(R1) define how the relations of syntactic dependence,

hierarchical by nature, are transformed into a graph. These

rules define the topology of the graph through two

components: the first one is a syntactic dependence

structure, represented as a tree, and the second one is its

corresponding representation in a graph. The rules match

sub-trees in the dependence trees of each sentence of the

text, and every time a sub-tree occurs, the words and

relationships defined by the rules, are added to the

resulting graph.

4. At this point, it is possible to extract words and text

relations, in order to make a representation showing how

words are related as concepts in the text. For this purpose,

the analysis is focused on the nouns, making them the only

words that will not be repeated in the graph, as a result

rules application. Assuming that words frequency and the

way they relate to one another reflects what the author

wishes to convey, the nodes and the arcs in the graph are

weighted, based on the number of times the rules match the

text. The visualization of the resulting graph, based on the

weight of the nodes (code of colors) and the arches

(thickness), allows the analyst to identify as concepts the

nouns that are jointly related in many sentences, since they

are represented by nodes with a high weight and having a

big number of incident edges.

Thus, the nodes of the morphologic structures graph are the

words of the text that occurred in at least one rule of

composition. The directed arcs represent the syntactic

dependence relationships found in the text, according to the

transformations defined by the rules applied during the

analysis.

5. The purpose of the third transformation is to focus on the

relevance of the relationships among the concepts, making

a semantic analysis based on the syntactic structure of the

text’s sentences. The suggested process contemplates

frontier and relevance analysis of nodes in the graph. The

analyst defines two parameters, one for the distance and

one for the relevance of nouns. The graph is explored

correlating only the nouns, to an extent given by the

distance parameter, and filtered preserve only those nouns

that have a significant number of relations, defined by the

relevance parameter. The distance defines the number of

arches between two nouns in the graph; the number of

relationships a noun has, defines its relevance in the

discourse.

The final result of the distance analysis is a non-directed

graph of text’s conceptual proximity approach (4). The

nodes represent the concepts in the text. The nodes’ weights

correspond to their ordinal organization. The arcs represent

the relationship between two concepts and arcs weight

represents the structural distance between two concepts. The

following figure (1) shows the general scheme of the

process:


Depending on the desired analysis, it is possible to define new

transformations, following a process that re

analysis, their application to the graph, and, finally that allows

visualizing the initial text

Designing the rule

The linguistic

of the sentences contained in the written texts. According to

Covington

analysis has been invented more than once.

such analysis, the words and the sentences are: “positioned

within categories as verbs and nouns, and those categories are

used in abstract rules, for instance, those of the N (noun) type,

NP (noun, pronoun), or VP (verb, pronoun)”

sentence will consist

verb

to a recurrent procedure permits the extraction of

word

mechanis

subject (observer) and the distinctions he makes. The principle

underlying the linguistic level theory relies in the finite

characteristic of the sentences and the principle of contention.

Even though the se

structures, a sentence could neither have infinite words nor lack

certain basic structures such as the

Although, the expression of an idea as a

several

of

where the higher levels become language rules.

A rule is understood as a sub

like

be

corresponds to a verb, thus match

associated with a VI label matches all verbs in infinitive.

apply the rule

order to define the

The

equivalent structures





Designing the rule

The linguistic


Covington







verb.


word

mechanis




Even though the se




several

of them, showing a hierarchic order on the sentence structure,



like labeled nodes

be generic:



apply the rule

order to define the

The set


4.





Designing the rule

The linguistic


Covington







. Consequently, the number of categories or rules associated


words from the text of an

mechanis




Even though the se




several internal structures, the s

them, showing a hierarchic order on the sentence structure,



labeled nodes

generic:



apply the rule

order to define the

set of


Figure 1. The general scheme of the process

APPLYING





Designing the rule

The linguistic


Covington [11]







Consequently, the number of categories or rules associated


from the text of an

mechanism




Even though the se




internal structures, the s




labeled nodes

generic:



apply the rule

order to define the

of defined



APPLYING





Designing the rule

The linguistic dependencies


[11], the linguistic and grammatical dependence









from the text of an

m to observe




Even though the se








labeled nodes

it may contain for instance the letter V, which



apply the rule, each

order to define the

defined



APPLYING





Designing the rules in grammatical dependencies

dependencies


, the linguistic and grammatical dependence









from the text of an

to observe




Even though the se








labeled nodes. This association with labels




, each

order to define the equivalent structure associated to the rule.

defined rules is one of the entries to the association of


Figure 2


APPLYING PAST AND





s in grammatical dependencies

dependencies











from the text of an

to observe




Even though the sentences may have different syntactic








. This association with labels




, each node of

equivalent structure associated to the rule.

rules is one of the entries to the association of

equivalent structures module

Figure 2


PAST AND




visualizing the initial text as


dependencies








sentence will consist of a noun sentence and a



from the text of an

to observe the analog relationship between the




ntences may have different syntactic












node of



module

Figure 2. A rule conformation


PAST AND




a graph.


dependencies analysis is based on the structures








of a noun sentence and a



from the text of an author. These networks offer a

the analog relationship between the








internal structures, the same principle is associated to



A rule is understood as a sub-tree of dependencies,





node of its



module of PAST.

A rule conformation


PAST AND




a graph.


analysis is based on the structures











author. These networks offer a









ame principle is associated to



tree of dependencies,



corresponds to a verb, thus matching


its sub



PAST.

A rule conformation


PAST AND THE




a graph.




















certain basic structures such as the noun or verbal clause.








ing all


sub-tree has a numeric ID,



PAST.

A rule conformation


THE ISSUED RESULTS


transformations, following a process that re-






analysis has been invented more than once. In order to carry out















noun or verbal clause.








all verbs


tree has a numeric ID,



A rule conformation


ISSUED RESULTS


-defines the rules of






In order to carry out
















sentence may have







verbs. Likewise, a node





A rule conformation


ISSUED RESULTS


defines the rules of










NP (noun, pronoun), or VP (verb, pronoun)” [19]



to a recurrent procedure permits the extraction of networks of









sentence may have




. This association with labels in the reul


. Likewise, a node






ISSUED RESULTS











]. That is



networks of









sentence may have



tree of dependencies, with

in the reul


. Likewise, a node





ISSUED RESULTS











. That is



networks of









sentence may have



with eagle

in the reul may


. Likewise, a node





ISSUED RESULTS











. That is, a



networks of









sentence may have

ame principle is associated to all


eagle-

may


. Likewise, a node

associated with a VI label matches all verbs in infinitive. To

tree has a numeric ID, in













, a



networks of









sentence may have

all


-

may


. Likewise, a node

To

in



Figure 2

the structural relationship between two subjects (N) and a

preposition (S), taking as structural equivalence the relationship

N

equivalence would be a gr

player

the context of CC Course.

The roles associated to the users are those of the supervisor

(professor), the analyst and the au

defin

empirical nature of the

development of the application in the CC course, rules in Table

1 were used; nevertheless, the exercise is susceptible to their

partial re

In the first

associated with

representation

the hierarchic

two subjects,

between

A first conceptual graph

In the grammar realm, it was established that grammatical

dependence comp

elaborate a text. As a result of using the syntactic analysis

platform, a response file is produced in terms of a graph. The

graph structure reflects the grammatical dependencies

words based on the definition of

Figure 2



N->N. An example is "chess player,"


player




defin




partial re

In the first

associated with

representation

the hierarchic

two subjects,

between



dependence comp





Figure 2



N. An example is "chess player,"


player->chess.




defining




partial re

In the first

associated with

representation

the hierarchic

two subjects,

between



dependence comp





Figure 2 shoews a





>chess.




the rules, unwanted results may arise due




partial redefinition.

In the first column of the table 1 represents

associated with

representation

the hierarchic

two subjects,

the first subject



dependence comp





shoews a





>chess. This document shows some of the rules used in








definition.

column of the table 1 represents

associated with a

representation as a

the hierarchical stru

two subjects, whose graph representation shows

the first subject



dependence comp





shoews an example of a rule. This rule contemplates





This document shows some of the rules used in








definition.

Table 1. Dependence and rule


a rule, and the

as a graph. Considering

structure of dependence associates

whose graph representation shows

the first subject



dependence component reflects the connections used to





n example of a rule. This rule contemplates










empirical nature of the definition process. For





rule, and the

graph. Considering

cture of dependence associates


the first subject



onent reflects the connections used to









equivalence would be a graph conformed by the relationship






definition process. For





rule, and the

graph. Considering



the first subject and












aph conformed by the relationship










rule, and the second

graph. Considering



and the second one.














(professor), the analyst and the author of the text himself. While







second

graph. Considering



the second one.






words based on the definition of structures. These structures








thor of the text himself. While







second column

graph. Considering the



the second one.






structures. These structures















column

the fi



the second one.










N. An example is "chess player," which structural






definition process. For the experimental




the dependence tree

column the

first row of Table 1,



the second one.










which structural






the experimental



the dependence tree

the corresponding

st row of Table 1,


whose graph representation shows the relationship










which structural






the experimental



the dependence tree

corresponding

st row of Table 1,

a verb with

the relationship










which structural





the rules, unwanted results may arise due to the

the experimental



the dependence tree

corresponding

st row of Table 1,

a verb with

the relationship





graph structure reflects the grammatical dependencies among





which structural





to the

the experimental



the dependence tree

corresponding

st row of Table 1,

a verb with

the relationship





among





which structural





to the

the experimental



the dependence tree

corresponding

st row of Table 1,

a verb with

the relationship





among



facilitate the understanding of the

by means of

trees. The following graph

to

introducing the methodology

Figure3. Dependence tree of a sentence and its corresponding graph

In terms of labels, we can observe the identification of the word

“game” in the dependence tree, which is "NCMS000",

indicating

(0)

the resulting gr

directed to the word tool. This means that the root

pointed to the dependent

the original text that reads:

the syntactic dependence

graph structure makes an analogue representation of the

grammatical dependencies

sentence

representations are also visi

the previous sentence, the arc

word tool is thicker than the arc

word incorporate. The number in each node

number of

word game appears with a

indicating that it matches three

user.


by means of


to the dependence tree of the se





indicating

(0) and no function assigned (0).

the resulting gr







sentence





number of



user.


by means of


the dependence tree of the se





indicating

and no function assigned (0).

the resulting gr







sentence





number of



user.


by means of







indicating a name (N), common gender (C), no degree assigned


the resulting gr







sentence (the relationships of the dependence tree). Such





number of matched rules including that




by means of the relationships







a name (N), common gender (C), no degree assigned


the resulting graph, the w







(the relationships of the dependence tree). Such





matched rules including that



Figure 4a.


the relationships









aph, the w















Figure 4a.


the relationships









aph, the w















Figure 4a.


the relationships

trees. The following graph (3)








aph, the word game is coupled with an arc


pointed to the dependent tool


the syntactic dependence relationships within the text. The


grammatical dependencies among









Figure 4a. Graph


the relationships

(3) is the result of

the dependence tree of the sentence: “

introducing the methodology” (originally in Spanish)






ord game is coupled with an arc


tool. This relationship corresponds

game is the tool,

relationships within the text. The


among


representations are also visible in the thicknes

the previous sentence, the arc joining the word g




word game appears with a three as number of occurrences


Graph

facilitate the understanding of the implicit

the relationships in

is the result of

ntence: “

” (originally in Spanish)





and no function assigned (0). This



. This relationship corresponds

game is the tool,



among the words contained in the


ble in the thicknes

joining the word g

word tool is thicker than the arc joining the word tool



three as number of occurrences

of the rules established by the

Graph for

implicit

the

is the result of

ntence: “






This example shows how, in




game is the tool,



the words contained in the


ble in the thicknes

joining the word g

joining the word tool





a paragraph

implicit semantics in

the associated

is the result of

ntence: “the game is the tool for






example shows how, in




game is the tool,





ble in the thicknes

joining the word g



matched rules including that word. For instance, the



a paragraph

semantics in

associated

is the result of applying

the game is the tool for










game is the tool, thus representing





ble in the thickness of the arcs

joining the word g


word incorporate. The number in each node corresponds to the

word. For instance, the



a paragraph

semantics in

associated

applying











thus representing





s of the arcs

joining the word game with the


corresponds to the




a paragraph

semantics in the text,

associated dependence

applying t


” (originally in Spanish).







directed to the word tool. This means that the root game


thus representing





s of the arcs

ame with the


corresponds to the




the text,

dependence

the rules








game


thus representing





s of the arcs

ame with the

joining the word tool and

corresponds to the




the text,

dependence

he rules








game is

. This relationship corresponds to

thus representing





s of the arcs; in

ame with the

and the

corresponds to the


three as number of occurrences,


the text,

dependence

he rules







is

to

thus representing





; in

ame with the

the

corresponds to the


,


F

the paragraph:

the acquisition and connection of different concepts and that

concepts are corporally interiorized. In the case of the course,

the learning and interiorizing processes occur

exercise of reading and writing essays related to those

readings, as well as designing games and in playing games

The distinctions made in the language compose the action

domin

action dominions of

interiorizing, readings, essays, and games. At the same time, we

can observe how the

acti

into a graph, and

arise that were not visible when the sentence

Besides the arc

the node

In the paragraph, the graph node will be the words: game,

design, readings, essays, exercise, interiorizing, learning,

acquisition, connection, concepts, pro

These words appeared in the paragraph composition at least

once. The syntactic dependence relationships of the paragr

are represented by the arc

the analysis.

A comparison of the conceptual graphs in three different

moments of the course (first, second, and third argumentative

essays) enables us to observe the characterization of his/her

discourse in the process of wondering who is

obse

Figure 4a and

the paragraph:






Figure 4b.


domin

action dominions of


can observe how the

action dominions. In the process of transforming a paragraph

into a graph, and


Besides the arc

the node







the analysis.





observer.

igure 4a and

the paragraph:






Figure 4b.


dominions. Based on the number of occurrences in the t

action dominions of


can observe how the

on dominions. In the process of transforming a paragraph

into a graph, and


Besides the arc

the nodes







the analysis.

5.





rver.

igure 4a and

the paragraph:






Figure 4b.


ions. Based on the number of occurrences in the t

action dominions of


can observe how the


into a graph, and


Besides the arc

s associated with nouns







the analysis.

APPLYING





rver.

igure 4a and figure 4b

the paragraph: “






Figure 4b. Corresponding conceptual graph of the graph 4a



action dominions of


can observe how the


into a graph, and


Besides the arcs

associated with nouns







the analysis.

APPLYING





figure 4b

“In the CC course, we consider learning to be






Corresponding conceptual graph of the graph 4a



action dominions of


can observe how the


into a graph, and the la


s thickness








APPLYING





figure 4b

In the CC course, we consider learning to be









action dominions of previous paragraphs are learning, concepts,


can observe how the paragraph’s


the later into a conceptual graph, new elements


thickness








APPLYING PAST IN DIFFERENT MO





show










previous paragraphs are learning, concepts,


paragraph’s


ter into a conceptual graph, new elements


thickness, the







are represented by the arcs according to the rules defined during

PAST IN DIFFERENT MO





show the result












paragraph’s




, their







s according to the rules defined during






the result












paragraph’s conceptual graph reflects these




colors represent the weights of

associated with nouns joined












the result












conceptual graph reflects these





joined in numerous



acquisition, connection, concepts, process, case, course, CC.









the results of applying PAST to

















in numerous



cess, case, course, CC.









of applying PAST to

















in numerous











discourse in the process of wondering who is his/her

of applying PAST to




the learning and interiorizing processes occur in the recurrent











arise that were not visible when the sentence was analyzed.


in numerous











his/her

of applying PAST to




in the recurrent











was analyzed.


sentences.







PAST IN DIFFERENT MOMEN




his/her observer as

of applying PAST to




in the recurrent


readings, as well as designing games and in playing games”



ions. Based on the number of occurrences in the text, the






was analyzed.


sentences.







ENTS




observer as

of applying PAST to




in the recurrent


”



ext, the






was analyzed.


sentences.





once. The syntactic dependence relationships of the paragraph


TS




observer as

of applying PAST to




in the recurrent



ext, the






was analyzed.






aph





observer as


As the distinctions in the paragraph of figure 4 emerge

conceptual graph in time t

others, the distinctions: emotion, observer, individual,

dispositions. In addition to the distinctions made by the author,

the conceptual graph relates the distinctions found. Generally

speaking, at this moment students cannot see clea

differences between the distinctions they make with the topics

of the course and the readings, and concepts and personal

appreciations.

The evolution of the exercise in time

appearance of new w

relations.

recursion, which: “makes reference to the application of an

operation on the result of the application of the operation”

A first recursion for the observer (student) starts to show when

th

to that of t2. Besides the distinctions existing in figure 3, in time

t2, the distinction “game” stands out.

Similarly, the evolution of the exercise in time t3 (figure 6)

evidences the appearance of new words, distinctions and

relations; the first idea that arises is how the strengthening of

the analysis is proportional to the increasing number of words

and their relationships as used within the text, just as is the case

in the conceptual map. The differentiation between the most

relevant distinctions (the ones with a higher number of

relations), and between the concept and the observer’s

appreciations justify the relations shown in the conceptual

graph. At this moment, the

how the author’s written discourse reflects the text: "so for me

to say that there is recursion, that there is language, in the case

of coordination of action, I must be able to make reference to

history”

analogically represented in a graph

observer to go back to the action dominions in any given

2

Figure 5.









appreciations.


appearance of new w

relations.




the essay in time t1 is taken as historical reference, as compared








n the conceptual map. The differentiation between the most








history”



A conceptual graph enlargement can be see

Figure 5.









appreciations.


appearance of new w

relations.




e essay in time t1 is taken as historical reference, as compared
















history”




Figure 5. Conceptual graph of Subj









appreciations.


appearance of new w

relations. In the relations, students recur




















history” [6]. That is




Conceptual graph of Subj









appreciations.


appearance of new w

In the relations, students recur




















. That is














appearance of new w





















. That is














appearance of new words and distinctions as well as





















. That is, working on the history of the essay














ords and distinctions as well as

















graph. At this moment, the idea of recursion provides insight on




, working on the history of the essay





t1


conceptual graph in time t1 (Figure 5)

























idea of recursion provides insight on








Conceptual graph of Subject X of the CC course in time

1


(Figure 5)

































ect X of the CC course in time


(Figure 5)






























analogically represented in a graph,





(Figure 5)





























enables the observer’s





identifies, amongst







The evolution of the exercise in time t2 evidences the


to the concept of






















A conceptual graph enlargement can be seen at



identifies, amongst







2 evidences the


to the concept of






















at appendix A



identifies, amongst







2 evidences the


to the concept of






















appendix A


As the distinctions in the paragraph of figure 4 emerge, the

identifies, amongst




speaking, at this moment students cannot see clearly the



2 evidences the


to the concept of






















appendix A


, the

identifies, amongst




rly the



2 evidences the

new

to the concept of


operation on the result of the application of the operation” [5].




Similarly, the evolution of the exercise in time t3 (figure 6) 2
















appendix A.

, the

identifies, amongst




rly the



2 evidences the

new

to the concept of


.




2
















moment, following the process bi

lineal direction, as it regularly occur

Al

one case of the results in terms of the conceptual graph, an

elementary analysis of the

written in the 2007 CC course

distinctions follow the chronology of the course. The

frequencies and the use of distinctions in the discourse

increased, particularly in the transition from sec

essay.

In this document, we have examined the relationships of the

conceptual structures contain

distinctions

(student) from the CC course may be able to accompli

h

syntactic analysis platform, which provides the opportunity to

work with a graph as an analogical representation of the text

written by the author. Such observations reflect the distinctio

of the author’s own “world” and the platform enables the

visualization of non

they appear in his/her essays. Thus: "our understanding of a

universe like that does not derive from discovering its presence,

bu

Written discourse structures the author’s distin

interactions of linear

to observe bi

than the conventi

organization within the text’s domain, and it shows the different

relational structures of the distinctions made by a particular

author.



Figure 6

Although the document, due to space limitations, reflects just







essay.



distinctions


his/her essays








but from recalling what we did to produce it"



to observe bi

than the conventi



author.



Figure 6

though the document, due to space limitations, reflects just







essay.



distinctions


is/her essays








t from recalling what we did to produce it"



to observe bi

than the conventi



author.



Figure 6. Conceptual graph of Subject X from the CC course in










distinctions, and the recurrent observations that the observer


is/her essays











to observe bi

than the conventi





Conceptual graph of Subject X from the CC course in

6.










, and the recurrent observations that the observer


is/her essays. Recurrent observations are facilitated by the











to observe bi-directional causalities or causal schemes rather

than the conventi






6.












. Recurrent observations are facilitated by the





visualization of non-






directional causalities or causal schemes rather

than the conventional one. PAST gives to a text a conceptual






DISCUSSION OF RESULT

















-lineal relations in the author’s discourse as





interactions of linear causality: the use of PAST is an invitation


onal one. PAST gives to a text a conceptual









elementary analysis of the distinctions reported in all the texts














lineal relations in the author’s discourse as





causality: the use of PAST is an invitation








time




distinctions reported in all the texts
























moment, following the process bi-directionally and not in one

lineal direction, as it regularly occurs.


time t3





written in the 2007 CC course, found that the number of





conceptual structures contained in a given text, their


















directionally and not in one

s.






found that the number of





ed in a given text, their





































































t from recalling what we did to produce it" [4]









DISCUSSION OF RESULTS







increased, particularly in the transition from second to the third













[4].









S







ond to the third













Written discourse structures the author’s distinctions in














ond to the third




(student) from the CC course may be able to accomplish on









ctions in














ond to the third




sh on




written by the author. Such observations reflect the distinctions





ctions in













ond to the third




sh on




ns





ctions in







Some possibilities for the next phases of the project are: a)

Being the purpose of the CC course is to induce students into a

self reflection process of observing their own observing process,

it will be interesting to let them use PAST as a means to think

about their written speech. In other words, if the students are

required to produce a text and a metatext on the structure of the

former, this metatext can be produced using the results of PAST

applied to their texts. This could be used starting from the

second version of the essay. b) In a higher recursion level, it

could be interesting to encourage a discussion among the

students regarding the diverse semantic structures produced

individually about a common topic. In this case, PAST could

once again be used as a means for guiding these discussions. c)

Another possibility is to apply PAST the proposed

bibliogrqaphy of the course. This could also generate particular

discussions vis a vis about the way the students structure their

texts. The structuring way used by the authors could be used as

a reference in the course.

7. CONCLUSIONS (TEXT ANALYSIS, RELATION

OF PROXIMITY)

Observing the texts linearly does not mean watching the

observer’s observer, but observing the observation process

carried out by the observer. The recurrent procedure defined by

the rules allows some words (concepts or distinctions the author

makes) to connect to other words, creating networks that spread

in different directions on a given author’s text.

According to Maturana, for a social system to exist, the

recurrence in the interactions resulting from the behavioral

coordination of its members must be adapted. Inter-subjective

coordination of actions occurs in the same language the subjects

live in as observers. Thus, the operations of distinction mold

networks of observing-subjects.

The previous exercise shows how, from the perspective of

cybernetics, the tool makes possible to confront distinctions.

The distinctions made by students (observers) are much more

than the sum of their sentences’ meanings and the elements

forming their words. In the graph, the observer finds the origin

of an idea, providing a new meaning that will be assimilated in

the cognitive structure. In other words, a coordinated meaning

of his/her discourse expressed in their own words, which will

give the observer the opportunity to think of new meanings and

make new distinctions in the next iteration when re-writing their

previous text.

Conceptual graphs reflect the distinctions made by each student

(observer) each time the essay has to be submitted. In node

(unitary referents) and arches topology, distinctions only

articulate the associations made by the observer, that take place

in the language, in a particular moment (t). Two occurrences of

the same word can have different meanings, and in a new

recurrent operation, these meanings should be communicated

within a new network of elements.

Although the analysis of structures is presented for the purpose

of understanding the conceptual relations of texts generated

during the CC course, the application of the platform is not

limited to this course. On the contrary, PAST has proven to be a

generic tool, and its operation and results can be applied in a

wide variety of fields. Some examples are marketing,

consumer’s preferences in terms of experience, or the analysis

of the different entries on a patient’s medical history, making

them “compatible” and enrich then a health database.

8. REFERENCES

[1] C. Ogden, and I. Richards, (1923). The meaning of meaning.

Routledge & Kegan Paul Ltd, Londres. Reprinted as El

significado del significado: una investigación acerca de la

influencia del lenguaje sobre el pensamiento de la ciencia

simbólica, Traducción de Eduardo Prieto, Editorial Paidos,

1984.

[2] H. v. Foerster, Las Semillas de la Cibernética. Editorial

Gedisa, S.A., 1992, pp. 40, 191

[3] G. Bateson, Mind and nature: A necessary unity. Nueva

Cork: E. P. Dutton, 1979, pp. 60, 69-70.

[4] G. Spencer - Brown, Laws of form. Nueva York, Bantam

Books, 1973, p. v, 104.

[5] H. Maturana, Biología de la Cognición y Epistemologia.

Ediciones Universidad de La Frontera. Serie Ensayos, Chile,

1990, pp. 59, 61.

[6] H. Maturana, & H. v. Foerster, "Biología, Cibernética y

Comunicación", Seminario organizado por la Asociación

Sistémica de Buenos Aires, Argentina, 1988.

[7] J. Bermeo, and R. Zarama, Application of the “team

syntegrity” invention as a proposed learning methodology:

Cybernetics of cybernetics, a case study. In proceeding of the

International Conference on Engineering Education, 2005,

Vol.1, pp. 494-500

[8] L. v. Bertalanffy, Teoría General de los Sistemas, México,

Fondo de Cultura Económica, 1976, p. 32

[9] L. Bloomfield, (1933). Language. London, Allen & Unwin,

1979

[10] L. Wittgenstein. Investigaciones Filosóficas (PU). Distrito

Federal. Universidad Nacional Autónoma de México, 1993, cf.

130 p. 21

[11] M. A Covington, A fundamental algorithm for dependency

parsing. Proceedings, 39th Annual ACM Southeast

Conference, Athens, Georgia, 2001.

[12] M. Foucault, (1971). L’ordre du discours. Gallimard, Paris.

Reprinted as El orden del discurso, Traducción de Alberto

Gonzalez Troyano, Tusquest Editores, 1983, pp. 11, 24.

[13] M. Lacobelli, (1949). The Semantic Discipline. The

Modern Language Vol 33, pp. 16-22.

[14] N. Chomsky, (2002). On Nature and Language,

Cambridge, Cambridge University Press (trad. esp.: Sobre la

naturaleza y el lenguaje, Madrid, Cambridge University Press,

2003).

[15] N. Chomsky, (1965), Aspects of the Theory of Syntax.

Cambridge (MA), The MIT Press (trad. esp.: Aspectos de la

teoría de la sintaxis, Madrid, Aguilar, 1971).

[16] R. Hernández, J. Bermeo, R. Zarama, Diseño, desarrollo

y validación de una herramienta computacional para

observar atributos en textos escritos bajo una metodología

particular de aprendizaje. Presentado en el XXI Congreso

Chileno en Educación en Ingeniería, 2007

[17] S. Moriello y Fritz. W. (2006). La Realidad es una

compleja red de interconexiones. Tendencias 21. [On

line]http://www.tendencias21.net/index.php?action=article&id_

article=389106

[18] S. Beer, Beyond Dispute: The invention of team

Syntegrity. John Wiley & Son, Chichester. 1994, p. 229.

[19] T. Honkela, Learning o Understand General Aspects of

Using Self-Organizizing Maps in Natural Language

Procesing. In proceedings of the CASYS´97, Computing

Anticipatory Systems, Liège, Belgium, 1979, p. 2.


APPENDIX A

Figure 7. Enlargement of conceptual graph in figure 6.


platform for the analysis of written texts in terms of .../sci/pdfs/zs050dd.pdf · [16]. past...

Documents