Page 177 - DCAP208_Management Support Systems
P. 177

Management Support Systems




                    Notes          Self Assessment

                                   State True or False:
                                   11.  Text mining is more difficult for computers to do.
                                   12.  A process called data mining is different from the mining of text.

                                   10.4 Web Mining


                                   Web mining is a form of information harvesting that applies to data gathered from online
                                   sources. Data collection from sources across the Internet allows users to aggregate large volumes
                                   of information for analysis to make key business decisions in an online environment.

                                          Example: A researcher might use web mining to collect information regarding the use
                                   of specific keywords in web content.
                                   Alternatively, retailers and other marketing professionals use online data mining to spot trends
                                   in web traffic, the conversion of site visitors to buyers, and other web usage.
                                   In terms of gathering, sorting, and analyzing data, web mining mimics traditional data mining
                                   activities. Comparatively, web mining activities focus on web-based information, rather than a
                                   large cross section of information sources such as off-line computer databases, customer records,
                                   or hard copy accounting data, as typically occurs with traditional data mining. Focusing solely
                                   on data collection from online sources provides targeted analysis needed for online marketing
                                   strategies, website structure decisions, and similar electronic commerce-related decision making.
                                   Collecting data via web mining also provides the added benefit of a broad international
                                   demographic, since websites from all over the world are available to researchers and information
                                   gatherers.
                                   Professionally, web mining is divided into three specific categories: web structure mining,
                                   usage mining, and web content mining. Each area focuses on specific information such as the
                                   structure and hyperlinks of a particular website, server log information regarding visitor usage,
                                   and specific content available online.
                                   Website analytic software packages and services are a prime example of web usage mining,
                                   providing webmasters with information regarding visitor traffic, search results used, links
                                   clicked, and time spent interacting with specific pages. Structure mining, on the other hand,
                                   provides detailed information about a specific website’s internal structure, including hyperlinks,
                                   databases, and query functions.
                                   To the marketing professional, web mining offers a wealth of uses relative to marketing activities.
                                   Knowing how site visitors use a particular website, how competitors set up a competing site,
                                   and what content is already online is valuable information. Such information helps key decision
                                   makers craft a marketing strategy based on previously proven techniques and documented
                                   information.
                                   Colleges and universities also utilize web mining via software that verifies student papers are
                                   unique and not plagiarized. Using web content mining principles, such grading aides search the
                                   entirety of the Internet for like content. Instructors upload the text of a student document and
                                   then instruct the plagiarism software to check the Internet for similar phrases or copied text
                                   online. Results are often expressed as percentage of matching text. Links to any similar results
                                   are provided to allow instructors the ability to visit sites to determine if matches are indeed
                                   plagiarized.






          170                               LOVELY PROFESSIONAL UNIVERSITY
   172   173   174   175   176   177   178   179   180   181   182