[Description]: This assignment is designed to provide students a good opportunity to read some materials to understand the theory, methods and recent development in the area of business intelligence,...

Business Intelligence
Business Intelligence 2
Running Head: Business Intelligence
Title: Business Intelligence
Student Name and Id:
Course Name and Id:
University Affiliation
Date of submission: 18/8/2019
Authors Note
The current report is presented as part of the requirements to complete the course work.
Abstract
The report details a range of Short text mining procedures employed for social media based information extraction. The findings in the report are detailed from information collected from various articles. Each of these articles are evaluated for the sake of content context, technology discussed, applicability of the same in Business intelligence context as well about variance of the technology discussed in the article with other technologies as well. The findings are detailed under different headings and generalized conclusions are presented in the report.
Contents
2Abstract
4Introduction
4Overview of short text mining and applications
5Survey Details
10Conclusion
11References
Introduction
Broadly speaking, Business intelligence is an integrated environment comprising a range of tools and technologies used by corporations to made decisions. Business intelligence does work for the sake of making decisions based as part of the strategic decision making in the organizations. Business Intelligence is actually a framework containing the set of applications, databases, software as well as hardware for the sake of users to enable the information analysis and to facilitate right decisions as per the needs of the organization. New technologies and advancements realized in recent years in the domains of the data mining and the Machine learning have radically changed the profile of the business intelligence. Data mining will provide the insights into the data and the patterns in the data, based on information obtained from Data mining, machine learning will provide useful insights into the actual business progression studies. Data mining and Machine learning combinations can typically work for determination of insights into fraud analysis, reduce inventory costing, market analysis, Risk analysis etc. Identification of anomalies can enable determination of the fraud and incoherence in the data inputs, further anomalies can also work on to enable organizations to predict the happenings which can enable them to prepare for the eventuality strategically. The quality and accuracy of forecasts in business settings got drastically improved in several domains with the advent of revolutionary changes in ML and Data mining at present. The focus of discussion in the current report is about short text mining. This will work for both the supervised and unsupervised learning in the short text categorization process.
Overview of short text mining and applications
Short text mining is meant for to get comprehensive overview of range of insights from the data available in hand. Both supervised and unsupervised machine learning can be employed in short text mining context. Typical applications of the text mining like categorization, Entity extraction, and sentiment analysis are some of the several techniques employed for the sake of extracting knowledge hidden in the text content. Enterprise business intelligence and data mining do employ the applications of the text mining for extracting useful information from the unstructured and qualitative data. It is possible for data mining operations to get key insights into the data available online and it can be used for tracing cyber crimes and cyber criminals as well. Efficiency of customer care services can be improved using text mining, infact dependency on customer care analytics can be totally eliminated using the short text mining exercises. Applications like Cogito Intelligence platforms are being employed for the sake of the extraction of required content and required insights from large pool of data available online at present. Other applications include spam filtering, social media data analysis etc. The following part of the report discusses the short text mining techniques relevant for contemporary applications.
Survey Details
(i) Injadat, M., Salo, F., & Nassif, A. B. (2016). Data mining techniques in social media: A survey. Neurocomputing, 214, 654-670.
About the paper:
The author indicated that there are as high as 19 text mining tools commonly employed for the sake of data extraction from the social media sources since the recent times. However still three techniques are most commonly employed tools, SVM (Support vector machine), DT (Decision tress) and Bayesian Networks (BN) are more commonly employed strategies for text mining from the social media resources.
Technology of SVM
SVM or short vector machine is also called as short vector network and the applications of this technology mainly consist in classifying the data in the social media. It will do so by supervised machine learning. It works for regression analysis of all sorts of data viz., text, images and other miscellaneous data like hand written content. It is often called as non-binary linear classifier. The advantage of SVM is contained in its versatility and applicability. SVM classifies data into two classes.
SVM is reported to be one of the best techniques to face the challenges of the classification in general. The technique can work well when there is high dimensional feature space as well when there is smaller training data as well. The technique is quite suitable for offline clustering as well. However SVM technique does suffer with difficulties like sparse context links and associated problems. DT particular suitability rests in its capacity to find the random variables estimation and their significance in the classification aspects. DT is robust and can be applicable for range of applications in general. BN is efficient in text clustering operations; it is also a simple Classification algorithm with very low computation times.
Application of text mining (SVM) in Business analytics
SVM is classical and finding place in literature since 2008 till date, the versatility of the technique is unchallenged, there were several new technique evolving but the applicability of SVM...