INTRODUCTION
Without a doubt, the most precious value that the human race has is knowledge, which has been stored in different ways over the years.
Today this knowledge can be known by the name of data or information, which is no longer only stored, but also analyzed to obtain results.
This is how the concepts of data, text and sentiment mining were born, in an effort to forecast the future and facilitate decision-making.
The concepts of data mining, texts and feelings are topics that appear in today's society due to the existence of large amounts of information, this concept has been developed in different areas of daily life, such as education, health, biology, market analysis, telecommunications, just to name a few.
They are computerized processes that identify behavior patterns in order to make a forecast of future actions.
HISTORY OF DATABASES
Thanks to the specialization of individuals over the years, the need arose to protect the information that they produced, such as investigations or property titles, basic information on the population including death, birth and marriage certificates.
With the passage of time, humanity was not satisfied with the physical protection of said information, so they gradually developed electronic devices that will facilitate data management, as well as a large amount of storage.
During the 20th century, various technologies such as photoelectric, the appearance of magnetic tapes, the first computers such as UNIVAC, were recorded, but it was not until the end of the 1960s that an article was published defining the model of bases. data as such.
Today large databases capable of generating, collecting and processing large amounts of information are managed, these characteristics are gradually maximized, becoming stronger and stronger (Martínez, E., 2000)
BIG DATA WITHIN ORGANIZATIONS
Organizations use this term more frequently thanks to various factors or storage media such as the so-called "cloud", Big data uses tools such as: text mining, data mining and sentiment mining in order to facilitate the decision making, making an organization increasingly efficient.
Thanks to advanced technology, the size of storage has increased, as well as the costs generated by the use of these elements for the organization.
The main objectives of the organizations are to be more efficient and obtain higher returns, so they add structured or conventional techniques and unstructured or semi-structured techniques to the analysis of their data, such as the aforementioned mining companies, which acquire their information from photographs, audio, video, books, among others.
That they are able to provide conclusions from analyzed data as quickly as possible, all this in order to obtain a competitive advantage, the data that is generally handled in this sense are the following:
- Data of a traditional (transactional) company. Generic information on customers, suppliers, supplies, etc. Data generated by data sensors. Generally, they are sensors integrated into different equipment since their production, they are also called smart meters, including chipsSocial data. All the social networks offered by the internet are included, the best known being twitter or Facebook, and elements such as photography and audio. Data generated by mobility. Generally provided by a mobile technology team, which can be a Tablet, Smartphone, or through Skype (Joyanes, L., 2013).
DATA MINING
It is born thanks to the constant management of information technologies, -there are different authors who are in charge of describing it, so the most significant definitions that have been identified will be mentioned.
"Data mining can be initially defined as a process of discovering new and significant relationships, patterns and trends by examining large amounts of data" (Pérez, C. & Santín, D., 2008)
Data mining is considered as a combination of semi-automatic techniques of artificial intelligence, statistical analysis, databases and graphical visualization, to obtain information that is not explicitly represented in the data and that may be more useful than the data of departure. (Martínez, E., 2000)
OBJECTIVES OF DATA MINING
Due to the existence of enormous amounts of information and the great advance of the so-called information technologies, specialized techniques have been generated for its analysis and control.
These techniques propose to carry out a precise identification of the information contained, that is, not only in a textual way, otherwise they discover patterns and trends of information, through different tools such as: algorithms, neural networks, data analysis, to mention some (Pérez, C. & Santín, D., 2008)
Data and information management have gone through different phases throughout history, finally reaching data mining, so it is necessary to consider the following representation to provide a clearer idea about it:
OPERATION OF THE DATA MINING
This technique identifies and extracts patterns that adequately "understand" the data and information, establishing a behavior forecast.
A clear example of data mining is the so-called ETL (Extraction, Transformation, Load) within its phases is data reading, incorporation, transformation, maintenance and data planning, (Pérez, C. & Santín, D., 2008) this model is presented as follows:
TEXT MINING
This technique is based on acquiring or finding useful information on a certain item from a certain amount of information, generally found added in specialized software
This technique focuses on the discovery of patterns, trends and associations among the diversity of existing information. Text mining covers the following areas:
- Categorization of texts Classification of texts Generation of clusters Discovery of associations Detection of deviations Trend analysis Construction of summaries (Montes, M., s / f)
The text mining process is divided into two stages:
- The first called processing. Where a representation of the texts is made in the form of structures and another discovery call. Where the objectives to be discovered such as patterns or trends are described (Tan, 1999)
Generally, the systems that are considered as text mining have simple representations and identification of some considered key words, which facilitates analysis and interpretation, so it can be said that their representations are delimited by established trends.
TEXT MINING APPLICATIONS
Text mining is applicable to various areas, the most prominent are:
- Security aspects. It is used to monitor the different sources of information, as well as for a more detailed study of possible ciphertext. Biomedical aspects. It is a newly developed aspect of research where natural language processing, computing, bioinformatics, including computational linguistics are analyzed. The fact of analyzing the relationship that organizations have with customers is exposed, to avoid the loss of them Academic applications. In many cases, when publishing various articles, indexing is requested, this refers to the existence of a certain quantity and quality of sides, established in a very specific way.
FEELING MINING
The already known as sentiment mining is a derivation of text mining, in this case an attempt is made to identify the intentions of customers or users in different media, for example email, Facebook, including different search engines with those that the network has.
The basic process of data mining is done as follows:
- An attempt is made to understand or relate the text to a specific item, for example with a particular business, restaurants, or travel agencies, after which the intention of the client is identified, placing a label on the phrases that make it up as positive or negative, for example, if the word “great” is classified as good in the databases, a positive label is added, otherwise, if the word “poor” is classified as bad, a negative label is added.
Through the analysis of these labels the organization has greater knowledge about the opinions and afflictions of its consumers. (Pérez, S., 2016)
ETHICS IN THE USE OF MINING
The information that is handled in some databases comes from users, who on many occasions are not fully aware that different organizations will use it for the purposes they deem appropriate.
Electronic purchases, payments, telephone calls, among others, are information that is generated every day and that is properly recorded, society is not aware of the large amount of information that it generates and the dangers that they can cause (Garriga, A., 2004)
Thanks to this information, data mining shows the type of influence that can be established with society to obtain certain benefits.
However, not everything is against these mining operations, this technology supports organizations to make their employees more efficient, analyzing their behaviors and skills.
Even by tracking browsers or email, it is possible to inquire if a collaborator is in search of new job opportunities, or if he presents any conduct that harms the activities of the company itself, including terrorism. (Franganillo, J., 2010)
BIBLIOGRAPHIC REFERENCES
- Franganillo, J. (2010). Implicaciones éticas de la minería de datos. Anuario thinkEPI.Garriga, A. (2004). Tratamiento de datos personales y derechos fundamentales.Joyanes, L. (2013). Big Data: Análisis de los grandes volúmenes de datos en organizaciones (Primera). México D.F.: Alfaomega.Martínez, E. (2000). Minería de datos, una herramienta para la toma de decisiones. UNAM. Recuperado a partir de http://132.248.9.195/pd2001/287820/Index.htmlMontes, M. (s/f). Minería de texto: Un nuevo reto computacional. México D.F.: IPN. Recuperado a partir de http://ccc.inaoep.mx/~mmontesg/publicaciones/2001/MineriaTexto-md01.pdfPérez, C., & Santín, D. (2008). Minería de datos, técnicas y herramientas (Ediciones Paraninfo S.A.). España. Recuperado a partir de https://books.google.com.mx/books?id=wzD_8uPFCEC&printsec=frontcover&dq=miner%C3%ADa+de+datos&hl=es419&sa=X&sqi=2&redir_esc=y#v=onepage&q=miner%C3%ADa%20de%20datos&f=falsePérez, S. (2016). Análisis de sentimientos. Universidad Central de Venezuela.Tan. (1999). The state of the art and challenges, Proc. of the Workshop Knowledge Discovery from advanced Databases. text mining.
