Frequent Itemsets and it’s applications in data analytics

Frequent Itemsets :
One of the major families of techniques for distinguishing data is the discovery of Frequent Itemsets. The main problem is seldom viewed as discovery of “association rules”, whose discovery depends radically on the discovery of Frequent Itemsets.

Frequent Patterns :

• Frequent patterns are patterns ( for example, Itemsets, or substructures) that comes frequently in a data set.

  Example –
  For example, a set of items, such as bread and butter, appear frequently together in a transaction data set which is a frequent itemset.

• A substructure can allude to different structural forms, such as subtrees or sublattices, which may be combined with subsequences.
• If a substructure take place frequently, it is called a structured pattern.
• Finding frequent patterns plays a crucial role in mining associations, correlations, and many other innovative relationships among data.
• Furthermore, it helps in data classification, clustering and other data mining jobs.

Application of Frequent Itemsets :

• The native application of the market – basket model was the analysis of true market baskets. That is supermarkets and chain stores documents the contents of every market basket brought to the register for checking.
• Here “item” are the various products that the markets and store sell, and the “baskets” are the set of items in a single market basket. A major chain might sell 10, 000, 000 variety of items and collect data about billions of market baskets.
• By identifying frequent Itemsets, a retailer can figure out what is commonly bought together. The important part is pairs or larger sets of items that occur much more frequently than what would be looking for were the items bought separately.
• We can identify by this analysis that many people buy bread and butter together, but that is of little appeal since we already knew that there were popular items individually. We might discover that many people buy hot dogs and cheese together.
• This will not shock the people who buy hot dogs but it gives supermarkets a chance to earn good profits by clever marketing. They can put hot dogs in the sale and will elevate the price of cheese. When people will come to buy hot dogs they will necessarily buy cheese without considering its high price.
• One more example is beer and diapers . Through data analysis supermarket salesperson observed that people who buy diapers, probably will have a child in their house and if they buy diaper they will usually not buy beer, they will unlikely to be drinking in a bar.

Features :
Now, you will see the features of frequent Itemsets in data analytics. The same model can be used to mine a lot of other kinds of data Some examples are listed below.

1. Related concepts :
   Let us consider items be words and baskets be documents. If we look for sets of words to occur together in many documents, the sets will be ruled by the most common words.
2. Plagiarism :
   Let consider the items for documents and the baskets for sentences in case of plagiarism. A document is “in” a sentence if the sentence is in the document. This ranging appears backward but is exactly what we need. In this application, we look for a pair of items that occur together in several baskets. If we got that pair, then we are ready with two documents that share several sentences in common. In practice, even one or more sentences in common are a signal to plagiarism.
3. Biomakers :
   We for instance let consider biomakers as blood proteins and ailment. Every basket is a set of data about a patient and their genome and blood – chemistry analysis, as well as their medical past of ailment. A frequent itemset that consists of one disease and one or more biomakers recommend a test for ailment.