Data Mining and the Rough Set Approach

Data Mining, a sub-branch of computer science, involving statistics, methods and calculations to find patterns in large amount of data sets, and database systems. Generally, data mining is the process to examine data from different aspects and summarizing it into meaningful information. Data mining techniques depict actions and future trends, allowing any individual to make better and knowledge-driven decisions.[1][2]

Data mining can aid direct marketers by providing them with useful and accurate trends about

In its infancy, data mining was as limited as the hardware being used. Large amounts of data were difficult to analyze because the hardware simply could not handle it [1]. The term "data mining" first began appearing in the 1980 's largely within the research and computer science communities. In the 1990 's it was considered a subset of a process called Knowledge Discovery in Databases of KKD [1]. KKD analyzes data in the search for patterns that may not normally be recognized with the naked eye. Today however, data mining does not limit itself to databases,

Data, Data everywhere. It is a precious thing that will last longer than the systems. In this challenging world, there is a high demand to work efficiently without risk of losing any tiny information which might be very important in future. Hence there is need to create large volumes of data which needs to be stored and explored for future analysis. I am always fascinated to know how this large amount of data is handled, stored in databases and manipulated to extract useful information. A raw data is like an unpolished diamond, its value is known only after it is polished. Similarly, the value of data is understood only after a proper meaning is brought out of it, this is known as Data Mining.

With rapid advancements in the technology, new concepts are hitting the industry and it is redefining itself over a course of time. The data mining is one of its kind to improvise the lives of people. Data mining uses techniques which are helpful in finding out the different forms of data. The data mining is closely related to the database technology. Almost every industry takes the help of the datamining to grow in their respective fields. For instance, stock management, quality control, risk management, fraud detection, marketing and analysis of investments. It has its applications ranging from finding the molecule structure of the gene to identifying a robbery at an international level.

Nowadays, data mining and machine learning become rapidly growing topics in both industry and academic areas. Companies, government laborites and top universities are all contributing in knowledge discovery of pattern recognition, text categorization, data clustering, classification prediction and more. In general, data mining is the technique used to analyze data from multi perspectives and reveal the hidden gem behind the enormous amount of data. With the explosive growth of data collections, it becomes time-consuming less effective to extract valuable information from massive databases through the use of traditional data analysis methods. An alternative way to solve this problem is to apply data mining, given considerations

Opinionated text has created a new area in text analysis. Traditionally fact- and information-centric view of text is expanded to enable sentiment-aware applications. Now a day, opinion or sentiment extraction is very important task for both business and academic world. A producer would want to know what people say about its products on popular site (Aciar et al.2007). A corporation would be concerned about the review of product given by user and accordingly monitor the effectiveness of its advertising campaigns. Therefore, sentiment analysis is popular stream, which extracts sentiments and analyze it (Bing Liu, 2010; Bermingham et al. 2009; Chen, 2008; Cummins et al.2010).

Generally, data mining (sometimes called data or knowledge discovery) is the process of analyzing data from different perspectives and summarizing it into useful information - information that can be used to increase revenue, cuts costs, or both. Data mining software is one of a number of analytical tools for analyzing data. It allows users to analyze data from many different dimensions or angles, categorize it, and summarize the relationships identified. Technically, data mining is the process of finding correlations or patterns among dozens of fields in large relational databases.

This research paper highlight the importance and need of data mining in the age of electronic media where large amount of information and consolidated database is readily available. This seemingly useful information can unearth some mind-blowing statistics and predict the future trends with relative ease through use of data mining techniques which can benefit the businesses, start-ups, country and individual alike. However, since data mining is effective in bringing out patterns, alerts, correlation and association through complex algorithms and analysis, it has, over the past few decades proved to be a useful

Data preparation is one of the most significant and time-consuming phases of data mining projects (Steinbach et al., 2005; Han et al., 2006; Yau, 2011). The data needs to be prepared to be in an appropriate state for analysis, maintaining its representativeness of the real world but in a format that is appropriate for the data analysis tools. Therefore, data pre-processing techniques like data selection, data cleaning, constructing new data, integrating data and transformation of data were used in this case study.

Data mining generally is the process of analysing data from different perspectives and summarising it into useful information (Thuraisingham, 1999). It is also called the “Knowledge Discovery in Databases” process. It can be understand in the way of discovering interesting and useful patterns and relationships in large volumes of data. The overall goal of the data mining process is to extract information from a data set and transform it into an understandable structure for future use. (Han & Kamber, 2006)

The system crawls all the information (customer reviews) from the particular product and starts to extract the feature words. The mining classification is divided into three subtasks. First, POS tagging, most product feature in reviews contain nouns and phrases. POS enables to identify these tags words by word. The authors have used NLProcessor linguistic parser to split the sentences and produce POS tag for every individual word. Some pre-processing is performed to remove stop words to avoid big POS tag lists. Fuzzy matching was also used to deal with misspelling and word variants. Second, identifying frequent features, features that are particularly used by customers in many reviews. To put in context, words, or sentences that occur together in many reviews. The authors use association mining to list out these words. An apriori algorithm based association miner is used to identify these features/words. The gathered features cannot possibly be genuine product features. Pruning enables to remove the unwanted features/words that don’t match to product opinion or feature. The author uses two pruning methods and compares their efficiency. While using compactness pruning, it looks for features that is composed of two or more words and identifies if they make

Opinions play a crucial role in the decision making process. Analysis in the field of making decisions and setting policies has shown that sentiment analysis and Opinion mining has become increasingly important in the field of Information Retrieval and Web analysis. In the past years, the growth of user generated data in web forums, social networking sites and other social platforms is tremendous, which diverts our study towards mining the opinions on web. In this paper, we have presented a novel methodology to classify the tweets and the complete opinion mining system is explained on the basis of survey and analysis. A flowchart has been proposed in which the overall picture of classification of twitter data has been proposed and accuracy of the evaluation strategies by various supervised learning algorithms has been evaluated. Review data is collected for various product domains from micro blogging sites like twitter, face book.

Social opinion has been analysed using sentiment analysis (SA). This is basically a natural language processing (NLP) application that uses computational linguistics and text mining to identify text sentiments as positive, negative and neutral. This technique is known as emotional polarity analysis which is related to text mining field, opinion mining and review mining. In addition, to calculate sentiment score, the sentiment acquired from the text is compared to a dictionary in order to determine the strength of that sentiment. Studies on sentiment analysis focus on text written in English such as sentiment lexicons while applying this to other languages will cause domain adaptation problem [12]. Traditional text classification is different from sentiment classification. Traditional text classification refers to pre-defined class to determine a document’s category, and it gauges the theme of the text itself, while the main aim of the latter is to determine the attitudes and opinion through mining and analysing user interest or other subjective information [13]. Studies in sentiment analysis have found that pre-processing the data is the procedure of cleaning and adapting.

Sentiment analysis consists of various elements out of which lexicons are an integral part. These lexicons are nothing but the words which make up a sentence. These lexicons can also be known as sentiment lexicon or opinion lexicons. It is important that we classify the lexicons we have into positives or negatives in order to achieve review analysis. One of the methods for sentiment analysis and classification is done on the basis of these lexicons. There are two broad approaches for sentiment analysis. These are lexicon based approaches and machine learning based approaches.