1. Web Mining and its types
Presented by :-
Nevil Shah
015964975
2. Web Mining
โ Web Mining is the use of Data Mining techniques to discover and extract useful
piece of facts and patterns from data over web services.
โ Discovering useful information from World Wide Web and its usage patterns
โ Proper mining of sessions in log analysis allows a web server owner to gather
interesting patterns about the users.
4. Web Content Mining
โ โProcess of informationโ or knowledge discovery from contents from the millions
of sources across the web.
โ It usually mines Web Data contents like audio, video, text,metadata, hyperlinks
etc.
โ In other words, Web content mining is the process of collecting useful facts and
figures from the web content.
5. Web Structure Mining
โ Web structure mining is a tool used for finding the association between Web Pages
that are related to the data.
โ It generates structural summary about the website and Web Pages.
โ Example:- PageRank Algorithm used by Google to determine the rank of a
page.
6. Web Structure Mining
โ Few more examples are :
โ Categorizing the web pages and the related information @inter-domain level
โ Discovering the nature of a hierarchy of hyperlinks in a website.
7. Web Usage Mining
โ Discovering userโs โnavigation patternsโ over the Internet which is dependent on the
web log information saved in clients, proxy server etc
โ Prediction of user behaviour while the user interacts over the Web
โ It is a mechanism to discover important usage patterns from Web content to
understand and better serve the needs of web-based applications
9. โ This step comprises of 3 steps:
โ Data Preprocessing
โ Knowledge Discovery
โ Pattern Mining
10. Data Preprocessing
โ The amount of data collected over the Internet is unbelievably huge, hence,
Data Preprocessing is a necessity to improve the value of information and
make the evaluation process smooth.
โ As a primary step, the noisy data is eliminated and thereafter, further steps
are implemented.
11. โ Data Preprocessing mainly includes 3 processes :
โ Data Cleaning
โ User Identification
โ Session Identification
12. Pattern Discovery
โ In this stage, the actions performed by the users over the internet are closely
noticed and interpreted.
โ For this step, the pages over the internet that are visited frequently by the
client are noted.
โ Few methods which include in this stage are :
โ Frequent Itemset Mining
14. Pattern Analysis
โ The concluding step of the Web Usage mining is pattern analysis stage.
โ The aim of this process is
โ to remove the irrelative impressions
โ to find the noteworthy impressions from the result of the pattern discovery
process.
โ Analysis methodologies and tools are: query mechanism like SQL, OLAP, and
Visualization etc
15. Examples of web mining
โ People with salary more than 50k USD and age greater than 40 performs their
share trading online.
โ Users X , Y and Z access similar set of URLs regularly.
โ User A usually buys an electronic product from this website at least thrice a
month.