1. Introduction
Day by day our world is invaded by an endless amount of information that arises with new technologies and research, since it is inevitable that through an analysis or treatment of information new information is generated.
But among all that information, there is a large percentage of it that cannot be quantified as it lacks precision or accuracy; since this is characterized by being so ambiguous.
Therefore, it is normal for an organization to have inaccurate information or better known as fuzzy information in its databases; since the human being is accustomed to including in his language terms or adjectives that cannot be measured; but that nevertheless are of great help to human beings during decision making.
But in such a fast world, it is impossible for the human being to analyze all the information of his organization to make adequate decisions in short times; therefore this analysis process has been replaced and can now be done by computers or software.
But for this process to develop, fuzzy logic or fuzzy set theory needs to be implemented. The following article will explain what is fuzzy logic ?.
2. Conceptual framework
2.1 Information
Information is a set of data about some event, fact or phenomenon, that organized in a given context have their meaning, whose purpose may be to reduce uncertainty or increase knowledge about something (Thompson, 2008).
2.2 Fuzzy
It is an adjective that refers to the fact that it is unclear, confusing, inaccurate, and not concrete, as it lacks a good definition or certainty.
2.3 Fuzzy logic
Fuzzy logic plays a great role in the development and interpretation of fuzzy information, so it is necessary to know what is fuzzy logic? And where does it arise?
Fuzzy logic or better known as fuzzy logic, which in Spanish refers to fuzzy logic or heuristic logic, deals with knowledge or information that lacks accuracy, but this does not mean that because it lacks precision, the information must be discarded.
Well then we would enter into a great controversy, because the amount of diffuse information that human beings handle and analyze to make decisions in their daily lives is surprisingly exorbitant.
Fuzzy logic is a computational intelligence technique that tries to understand or analyze information with a high degree of imprecision.
By imprecise information, I mean information that is made up of expressions like "it is very hot", but it does not really allow us to know exactly how much heat is being experienced.
The key to understanding this type of expressions of daily use by the human being lies in the understanding of the quantifiers of our language, such as: "a lot", "very", "scarce", "few", among others.
Fuzzy logic differs from conventional logic in that the former processes imprecise information and the latter processes only well-defined and accurate information.
The most surprising thing is that the human being does understand the expressions that make the information imprecise and, based on them, makes important decisions.
However, the analysis of the quantifiers is carried out by humans, although the previous one cannot be carried out by computers.
This ended when the researcher Lofti A. Zadeh, not being satisfied with the options of the classic sets (crisp sets) that only have two options: belonging (1) or not belonging to a set (0).
He first described in 1965 the "fuzzy sets" or "fuzzy sets"; where it stands out that the sets can partially belong to some sets, leaving behind the traditional or conventional logic. This concept was published in the article fuzzy sets.
Zadeh establishes the concept of fuzzy logic, when he tests the theory of fuzzy sets. Six years later Zadeh publishes in the article “Quantitive Fuzzy Semantics”, where he defines the elements that make up fuzzy logic and its current applications.
It should be mentioned that the fuzzy set theory and the concept of fuzzy logic at the beginning had great acceptance in Europe, while in the United States (which was their cradle) they refused to accept this new discovery; since scientists were skeptical of anything other than binary theories.
But it was not until 1974, when Ebrahim Mandani, applied for the first time the concept of fuzzy logic in the field of control in the industry.
He was the creator of the fuzzy control or fuzzy control system; taking it to the applying it to regulate the steam engine.
It is to be imagined that Mandani's success in applying fuzzy logic and fuzzy sets was a consequence of technological advances in computers (which were not as well developed before).
3. Fuzzy information
Engineer Zadeh thought that people did not need logical or numerical information to perform control tasks; instead, it was sometimes much more practical for people to carry out activities or tasks with imprecise data.
Only then is fuzzy information that information that is made up of a set of fuzzy data or fuzzy or unclear data, which are not exact and are not quantifiable, however such data is qualitative and can be measurable.
It is extremely important to mention that diffuse information is of great importance for the development of organizations, since this can represent a limitation by blocking the organization from acquiring true knowledge when sharing or informing it.
An example of fuzzy or unclear information is when interrelated data is not quantifiable, an example is as follows:
During a market study questions should be asked to know the degree of acceptance that a product will have in consumers.
By means of the questions it is tried to know the emotions that the consumer perceives about the product, these questions are closely related to a series of organoleptic tests; which will result in a set of data that is not measurable in other words is "fuzzy information" that is not clear enough to be understood.
It should be mentioned that diffuse information is not only that which is not quantifiable, but can also be all that logical information that loses its true meaning during its processing or analysis.
The fact that we have diffuse information does not mean that it cannot be used or that it has no value for the organization; on the contrary, it means that we have to subject said data to another process in order to analyze it and transform it into logical, conventional or traditional information.
Once the information has been transformed into quantifiable information, it will provide the necessary tools for the organization or individuals to make good decisions.
3.1 Fuzzy set
The father of the fuzzy logic Zedah, defines the fuzzy set as those items that the degree of belonging of an item to a set is not known for certain, but which are related in some way.
An example is the set of tall people, this is a fuzzy item; Because there are no measurable parameters that allow us to identify or know from what extent people can be considered tall so that they belong to the group of tall people.
So it can be said that the height limit is diffuse and since there are no specifications or attributes, anyone can enter that set.
A fuzzy set is defined as follows: A fuzzy set A over a universe of discourse U (ordered) is a set of pairs represented by a mathematical equation:
A = {µλ (u) / u: u ε U, µλ (u) ε}, Where:
µ = membership function
µλ = degree of membership of element u to the fuzzy set A.
(ranging from 0-1)
µλ (u) = 0 indicates that u does not belong at all to fuzzy set A.
µλ (u) = 1 indicates that u belongs fully to the fuzzy set A
3.2 Elements of the fuzzy set
Once we explain what is and the importance of the diffuse information in the organization, it is necessary that we know the elements that the diffuse information has.
So an information item is made up of (Urrutia & Varas, 2013):
Á Attributes: is a function that determines the value of an object.
Attributes are classified into three types:
Attribute classification
Type I: Refers to accurate or better-known classical data, which may have defined linguistic labels. Example that of a tall person.
Type II: attributes that store classic and fuzzy data, being an imprecise data set related to an ordered referential.
For example the labels can be: child, youth, adult with reference to a set of 0 to 100.
Type III: they are attributes on imprecise data related to an unordered normal referential.
Example: the attribute would be hair color and the referential not ordered or the labels would be; blond, redhead and brown.
The aforementioned attributes are used to make a model to deal with fuzzy sets.
- Object: it can be anything or item.Value: it is a subset of the reference domain associated with the attribute (Garrido & Cadenas, 2013). Confidence: it is an indicator of the degree of veracity of the item.
3.3 Characteristics of fuzzy information
Fuzzy information is characterized by being imprecise and having a high degree of uncertainty, the difference between these two characteristics is that the imprecision is related to the item value and the uncertainty is an indicator of item confidence.
3.4 How to recognize the diffuse information of the conventional one?
It is normal that during the information gathering stage all kinds of information are obtained, both logical and diffuse, since an organization generates a quantity of information in its different departments and environment; external and internal.
To determine if the information obtained during the collection process is diffuse, it is necessary to consider a series of questions considered by Rosas Sánchez (2012):
- Does the information obtained satisfy the needs of the organization? Can the information be consulted in an efficient way by the collaborators who require it? Is the information concrete, reliable and fully integrated? Does it present any duplication?
3.5 Tools for the identification and extraction of fuzzy information
The sequence of steps that will be described below are applied for the identification and extraction of the fuzzy information from the conventional one, in order to obtain only the fuzzy information and then the fuzzy information can be processed and worked to transform it into logical information..
- Assign personnel to analyze the pertinent information of each activity. Consult expert personnel in the development of a method for information management in the organization. Implement a method to collect, collect all the information and structure and order it. Hierarchically. Determine the sources of information that are being handled. Measure the reliability of the sources of information. Implement rapid reading and selection techniques to underline the main ideas, within which the "key words" must be identified. content.Use thematic indexes and search engines.
3.6 Tools to analyze fuzzy information
Once the fuzzy information has been identified and extracted, it is necessary to implement other types of tools to carry out the analysis or the conversion of fuzzy information into conventional information; this in turn becoming knowledge for the organization.
3.6.1 Analysis of quantitative data
The quantitative information is generally not diffuse, due to the degree of precision and reliability it possesses and since they are measurable; What has to be done to prevent the quantitative information from becoming diffuse is to implement software that facilitates the manipulation of the numerical data sets, in order to maintain their reliability and certainty for decision-making.
3.6.2 Analysis of qualitative data
The analysis of these data is more complicated, due to the lack of certainty, reliability and since the data of the fuzzy set does not have a limited structure or belonging.
In other words, the qualitative data are not measurable, so they complicate their analysis for study, being only subjective concepts, their knowledge becoming diffuse.
3.6.3 Heuristic methods
Heuristic methods are used to implement a creative method for solving diffuse problems, among which is the Silver meal method, heuristics provide better solutions when executed by software, obtaining the ideal solution.
3.6.4 Fuzzy logic
It is one of the newest methods to analyze qualitative or imprecise information. This method is widely used when a concept has different meanings depending on time and context. The basis of fuzzy logic is the fuzzy sets by Zadeh mentioned above.
3.7 Model of a system based on fuzzy logic techniques
The model of the system based on fuzzy logic techniques is made up of three blocks (see figure 2):
Diffusion
Inference
Broadcast
Fuzzy logic systems
Figure 2 shows a system which starts with the data entry that is made up of the set of variables, which were previously selected according to the specifications of the problem to be solved. An example would be the factors related to transportation in an urban space: population, number of vehicles, homes, employment.
3.7.1 Diffuser block
It has the function of forming fuzzy sets, according to the degree of belonging to each variable of the input (individually entered) it will be decided in which fuzzy set they will be placed.
3.7.2 Inference block
It has the function of relating or indicating the fulfillment of one proposition with another, which is represented pq, in this way it can be expressed if… then.
Therefore the inference block expresses fuzzy rules like If u is A then v is B, where A and B are fuzzy sets; where u ε U and v ε V.
The final objective is to transform the fuzzy sets formed by the previous block according to the degree of membership into fuzzy sets associated with an output variable.
3.7.3 Broadcast Block
Once the fuzzy sets were associated with an output variable, it is necessary to specifically express the results obtained.
It is important to mention that this stage can be carried out using three methods:
Cut to: consists of a reduction of the total values to a single set.
Maximum value: the set only has a maximum value, in case there were different maximum values, it is necessary to take the values from the averages.
Centroid: center of gravity, calculated with summations.
Outputs: these are the concrete results, which in the example would be knowing or identifying vehicular traffic with respect to the number of urban trips.
3.8 Communication of fuzzy information
Communication is essential in any process within the organization and the communication of fuzzy information is not the exception, since it is necessary to find a way to avoid that the organization has more fuzzy information than logic.
For this reason, it is recommended that once the organization analyzes and quantifies the diffuse information, the collaborators who are positively or negatively affected are informed of the specific results obtained at the end.
3.9 Example of fuzzy set theory
A basketball coach must select candidates for his team, the specifications or parameters that players must meet to belong to the selection are: height ≥1.85 m and score ≥13 baskets out of 16 (see Table 1).
Table 1. Results obtained from the test (D´Negri & de Vito, 2006).
Results obtained from the test
The classic solution would only take candidates I and F; However, candidate E scored 16 of 16, but not having the stature limit is out of the solution.
The resulting solution for applying fuzzy sets is different, you must first define fuzzy numbers for each variable ranging from 0 to 1 according to the degree of belonging to the set of height ≥1.85 m and score ≥13 baskets of 16, obtaining at the end a combination of diffuse logic.
So applying set theory we can realize that we get another solution; where player E is part of the final solution, while the other classic solution would be a good element for 2 cm.
Table 2. Solution by fuzzy logic (D´Negri & de Vito, 2006)
Fuzzy logic solution
The difference between both methods consists of graduation, while in the classical solution the graduations are 0 or 1 (belongs or does not belong), the solution for fuzzy set theory the graduation is intermediate between 0 to 1; expanding the degree of membership in the fuzzy set (see table 3).
Table 3. Comparison of the graduation scale obtained through classical logic and fuzzy logic (D´Negri & de Vito, 2006).
Comparison of the graduation scale obtained through classical logic and fuzzy logic
3.9.1 Examples of the application of fuzzy information in the industry applying fuzzy logic.
Japan is the country that has implemented fuzzy logic for solving fuzzy information problems, listed below.
- Air conditioning control systems. Autofocus system in cameras. Optimization of industrial control systems. Handwriting recognition system. Efficient improvement in the use of fuel. Simulation of human behavior. Store and consult imprecise information (FSQL). Fuzzy logic by NASA to perform complex maneuvers Maintenance systems: For infrastructure that is related to failure or maintenance; Call them bridges, metal structures, roads, control systems.
Control systems (decisions for regulation, for example in traffic, water tanks, floodgates).
Fuzzy input calculations (loads on tracks and bridges, seismic stresses, slope stability).
In medicine
- Evaluation of arterial aging by means of artery signals, using bloodless registries. Registry classification to evaluate institutional costs of interned diabetics. Fuzzy techniques for the analysis of biomedical data.
4. Conclusion
As on other occasions I have already mentioned that information is the most important asset within organizations and is the asset that defines the degree of competitive advantage that each organization possesses.
But this does not mean that the organization simply by having a large amount of information will have a competitive advantage over its other competitors.
Well, it is useless to have information that cannot measure the impact that a certain variable can cause or in the same way it is useless to have a large amount of conventional information if you did not carry out the correct process and therefore you do not have a high degree of confidence or in the information that is supposed to be one hundred percent reliable, since it causes the collaborators of the organization to be lost among the seas of information that it has.
Then the organization can possess two types of information; logical or conventional information (which is quantifiable) and fuzzy information characterized by not being quantifiable and by being information that lacks precision.
However, diffuse information is of great importance for the organization, so it is necessary that said information undergo a process of transforming diffuse information into concrete information that serves as knowledge to the organization and in turn allows it to make correct decisions.
The analysis of fuzzy information is based on fuzzy logic.
5. Thesis proposal:
Implement fuzzy logic to determine the causes that cause the most concurrent pipeline failures in PEMEX, Cd. Mendoza.
5.1 Objective:
Transform the diffuse information into conventional information that serves as knowledge for emergency personnel, in order to make production processes more efficient.
6. Bibliographic references
- Arriola Carrera, GA (2013). Obtained from http://es.scribd.com/doc/57256980/INFORMACION-DIFUSACancino Velásquez, JA (May 5, 2012). Communication and handling of diffuse information. Obtained from Gestiopolis: http://www.gestiopolis.com/administracion-estrategia-2/comunicacion-manejo-de-informacion-difusa.htmCivil, I. (April 28, 2013). Elements on fuzzy sets and fuzzy logic. D´Negri, CE, & de Vito, EL (2006). Introduction to approximate reasoning: fuzzy logic. Obtained from Argentine Journal of Respiratory Medicine: Garrido, & Cadenas, JM (April 28, 2013). Treatment of fuzzy information in a Gaussian component mixture model, Rosas Sánchez, L. (March 14, 2012). Handling of diffuse information. Indispensable tool for decision making. Obtained from Gestiopolis: http: // www.gestiopolis.com/administracion-estrategia-2/manejo-informacion-difusa-heráculo-indispensable-toma-decisiones.htmThompson, I. (2008). Promonegocios. Obtained from http://www.promonegocios.net/mercadotecnia/definicion-informacion.htmlUDLAP. (April 29, 2013). Diffuse logic. Obtained from Chapter 3: http://catarina.udlap.mx/u_dl_a/tales/documentos/lmt/ramirez_r_o/capitulo3.pdfUrrutia, A., & Varas, M. (April 28, 2013). Diffuse Data Model in UML: A Case of Advice for Advertising Insertion. Obtained frommx / u_dl_a / tales / documents / lmt / ramirez_r_o / capitulo3.pdfUrrutia, A., & Varas, M. (April 28, 2013). Diffuse Data Model in UML: A Case of Advice for Advertising Insertion. Obtained frommx / u_dl_a / tales / documents / lmt / ramirez_r_o / capitulo3.pdfUrrutia, A., & Varas, M. (April 28, 2013). Diffuse Data Model in UML: A Case of Advice for Advertising Insertion. Obtained from
