Essay:

Essay details:

  • Subject area(s): Marketing
  • Price: Free download
  • Published on: 14th September 2019
  • File format: Text
  • Number of pages: 2

Text preview of this essay:

This page is a preview - download the full version of this essay above.

This chapter refers to the diversity and importance of recommendation systems based on the need for management of large quantities of data that is accumulated from the users with the purpose of filtering the data after certain criteria. The purpose for which recomandation systems were made represent a particular object recommendation based on certain criteria and viewed from a certain point of view. Recommendation systems are very useful in certain situations having the speed of data filtering as the main advantage. These systems can be used in a variety of fields from recommending web pages, articles, parks, restaurants, objects that can be sold as cars , bicycles drinks, etc. At the same time these systems create a public profile based on different characteristics of the objects in question.

\\end{abstract}

}

%----------------------------------------------------------------------------------------

\\begin{document}

% Print the title

\\maketitle

\\section{Introduction}

\\lettrine[nindent=0em,lines=3]{E} -commerce sites can have big , essentially unlinked , catalogs or summary list of items . With large amount of information into catalogs comes the difficulty for buyers to use their standard search and browsing facilities. In a particular way in the case of normal buyers or occasional buyers and in the case of so complex products, the difference between  products specifications and the need understanding of the buyers can be so hard to bridge. A good e-commerce catalog must cover user needs to products that can fulfill them. This one way describes an interactive ,an incremental, a case-based  approach to solving this problem. The approach is

very interactive and incremental, so it doesnt require that the user have a completely

need at the start of his search. The system is based on each case that it emphasizes products over their features or their constraints, and uses case-based specified techniques for its product retrieval. The most common situation for a modern recommendation system is a website with which a user can interract.\\\\ \\\\ Tipically , a well working system presents a catalog of items to a user and the user should select the among of items to receive the list of details on an item or to interact with that item in a specific way. \\\\ \\newpage For example , a smartphone E-commerce website presents to the user the catalog in which the products have a small "story" full of descriptions and specifications near the picture of each item so the user can get and see more details about a selected product and also he has the oportunity to purchase the item in case.\\\\ \\\\   Although the web server will transmit the HTML and the user can see a web page, the web server typically has a database of all items and dynamically will construct web pages with a list of items. Because the fact there are often many more items available in a database than would easily can fit on a web page, it is necessary to select a "subset" of items to display to the user the items in case or to determine an order in which to items can be displayed.lettrine[nindent=0em,lines=3]{R} ecommendation systems use a number of different technologies . We can classify there systems into five broad groups.

\\begin{itemize}

\\item Collaborative Systems

\\end{itemize}

    

   This type of recommendation system can aggregate ratings or recommendations of the objects , can recognize commonalities between users based on their ratings , can generate new recommendations based in inter-user comparisons and can also use time-based discounting of ratings.

   \\begin{center}

    \\includegraphics[width=2.2in]{img/1.png}

   \\end{center}

Collaborative systems  have been applied to many different kinds of data including: sensing and monitoring data, such as in  environmental sensing over large areas or multiple sensors; financial data, such as financial service institutions that integrate many financial sources; or in electronic commerce and web applications where the focus is on user data, etc. This discussion is focusing on collaborative filtering for all the user data, although some of the methods may apply to the other major applications as well.

\\begin{itemize}

\\item Demographic Systems

\\end{itemize}

   The importance of this type of system is a big one because it is categorising users based on personal attributes, such as the age , habbits ,education , income , geographic place where they are living and it is making recommendations based on demographic classes (male, female).

   

   \\begin{center}

    \\includegraphics[width=2.5in]{img/2.png}

   \\end{center}

   

   It is well known the fact that if one item from a website catalog was purchased by a female this is not meaning that a male will do the same choice. This is the perfect reason for us to see the demographic impact on recommendation systems.I believe there is a good factor who influences the comparability which was receiving too little attention : users demographics and characteristics. For example let's say it is well known the fact that the results from one study cannot be used to make some conclusions for a population if the one who did the study does not belong to that population.\\\\

 For instance, in marketing you cannot make reliable conclusions about how elder people

in Germany will see a product (let's say the product is a smartphone) if a study about that product was conducted in France with students. They all will enjoy the apearence of the smartphone. Students will appreciate a set of characteristics of the smartphone such as is games are working, speakers functionality , etc , and the elder people will appreciate another set such as the signal quality or easy interaction between human and gadget. So demographic system has a big influence on taking a good decision when you want to purchase a product. That's why this system is an important recommendation system.

    \\begin{itemize}

    \\item Content-based Systems

    \\end{itemize}

   Content-besed recommendation systems have the effect of guiding the user in a personalized way for interesting objects in a big space of possible options. Content-based recommendation systems try to give some recommendations about items similary to those a given user has liked in the past. Indeed, the basic process which was performed by a content-based recommender consists in matching up the attributes of the user profile in which his preferences and his interests are stored , with the attributes of a content object or item in order to prepare a recommendation to the user new interesting objects or products.

   

   \\begin{center}

    \\includegraphics[width=2.5in]{img/3.png}

   \\end{center}   

   

   

   

   The abundance of the information which is available on the Web and in Digital Libraries, combined with their dynamic and heterogeneous nature, has determined a rapidly

increasing difficulty in finding what we want when we need it and in a manner which

well meets our requirements.As a consequence, the main role of user modeling and personalized information access is becoming so crucial: users need a personalized support in passing through large

amounts of valid available information, according to their interests and tastes.

The main problem of recommending items has been studied extensively, and two main paradigms have emerged. Content-based recommendation systems try to recommend items similar to those a given user has liked in the past, where as systems designed according to a collaborative recommendation paradigm identify users whose preferences are very similar to those of the given user and recommend items they have liked in the past. \\\\   

    Systems implementing a content-based recommendation approach are analyzing a set of documents and descriptions of the items previously rated by a user, and are building a model or profile of user interests based on the features of the objects rated by that user.

The profile is a structured representation of user interests, adopted to recommend new interesting items. The recommendation process basically consists in matching up the attributes of the user profile versus the attributes of a content object. The result are representing user s level of interest for that object.

   \\begin{center}

    \\includegraphics[width=2.5in]{img/5.jpg}

   \\end{center}

    The Architecture of Content-Based Recommendation system is well structured and need proper techniques for showing the items and producing the user profile and some good strategies for comparing the user made profile with the item representation. The Content-Based recommendation system process is structurated in 3 steps , each one with a new point of view.

    

    \\begin{itemize}

    \\item Content Analyzer

    \\end{itemize}

   When the information has no structure , a kind of pre-processing step is needed to extract the structured and relevant information.The main responsibility of this component is to represent the content of items (text,documents, Web pages, news, product descriptions, etc.) coming from the information sources in a form acceptable for the next processing steps. Data objects are analyzed by the extraction techniques in order to shift the object representation from the original information space to the target one . This representation is the input up to the PROFILE LEARNER and FILTERING COMPONENT;

   \\begin{itemize}

    \\item Profile Learner

    \\end{itemize}

   This module is collecting representative data of the user preferences and generalize this data, in order to build the user profile. Usually, this generalization strategy is realized through machine learning techniques, which are able to refer a model of user interests starting from objects liked or disliked in the past. The PROFILE LEARNER of a Web page recommender can implement a feedback method in which the learning way combines vectors of positive and negative examples into a new vector representing the user profile. Training examples are Websites on which the user provided a negative or a positive feedback.

    

    \\begin{itemize}

    \\item Filtering Component

    \\end{itemize}

   

   This module is exploiting the user profile to show relevant objects by matching the profile representation against that items to be recommended. \\\\

   \\ User Profiles \\\\

   

   A profile of the user is a profile full of his interests and is used by the most recommendation systems . This type of profile can have different types if information. There are 2 large categories of information contained by the user$'$s profile.

   \\begin{itemize}

    \\item[1-] User$'$s preferences \\\\

   

    User$'$s preferences are contained in a list full of description of the types of objects that user is interested of. There are many possible representations of this description, but one common representation is a main function that for any item description predicts the user$'$s interest. \\\\

    

    

    \\item[2-] User$'$s history \\\\

    

    A history of the user$'$s interactions with the recommendation system.This one stored the products that the user saw and the the other information about the user$'$s interaction (rating the product).

    

    

    \\end{itemize}

    \\begin{itemize}

    \\item Utility-based Systems

    \\end{itemize}

    

    Those systems make suggestions based on a calculation of the utility for each object for the user and also are employing constraint satifaction techniques to locate the best match for user$'$s preferences.

   \\begin{center}

    \\includegraphics[width=2.5in]{img/6.png}

   \\end{center}

Utility-based recommended system is attempting to model user$’$s multi-attribute utility function and is recommending objects with the highest utilities. The traditional utility--based recommended system needs that users rate items to extraction utility function, which made more user burden and lower satisfaction.

    \\begin{itemize}

    \\item Knowledge-based Systems

    \\end{itemize}

Knowledge-based recommendation systems perform a needed function in a world of always-expanding information resources. Unlike other recommender systems, they don$'$t depend of large bodies of statistical data about particular rated objects or particular users. \\\\ \\\\

Our experience has shown that the knowledge component of these systems need not be prohibitively large, since we need only enough knowledge to judge objects are as similar to each other. \\\\

 \\\\

Further, knowledge-based recommendation systems are actually helping users to explore and thereby understand an information space. Users are taking integral part of the knowledge discovery process, elaborating their information needs in the course of interacting with the system in cause.\\\\ \\\\ One need only have general mastered knowledge about the set of objects and only an informal knowledge of one$'$s needs and the system knows about the tradeoffs, category boundaries, and useful search strategies in the domain you searched on.\\\\ \\\\

Knowledge-based recommendation systems are strongly complementary to other types of recommendation systems just because of the functional knowledge:a particular item meets a particular need.

\\section{Recommendation systems in action}

\\lettrine[nindent=0em,lines=3]{W}here we see recomendation systems ? The answer is "everywhere" ! We daily meet websites where recommendation systems are playing an important role behind the scenes.\\\\ \\\\

A good example is Youtube. Youtube is quoted that there are roughly 100 hours of video recorded per minute.This is an example of information overload.There$'$s too much content which isn$'$t possible for us to see or find without somethig helping the content come to us. How do we still get what we like ? Simple. Recommendation systems.

   \\begin{center}

    \\includegraphics[width=2.5in]{img/8.jpg}

   \\end{center}

Another example where recommendation systems are very important. One very classic example is Netflix. Netflix is a website of movies or videos with a very well established recommendation system platform

   

   \\begin{center}

    \\includegraphics[width=2.5in]{img/9.jpg}

   \\end{center}

   

We met recommendation systems even on social networks. For example on Facebook. "People you may know" is one of the sections on Facebook where we have to deal with those systems.

   \\begin{center}

    \\includegraphics[width=2.5in]{img/7.png}

   \\end{center}

I left Amazon , well known as the best E-commerce website , at the last time because I want to talk about this website in a detailed way. This E-commerce website combines all Recommendation systems and a huge data provided by the user. The website is doing precise recommendations and it is able to suggest you things you never expected. It is combining art and science,typical fields of study around market analysis (called afinity analysis) which is a subset of the field of data mining.  A study made by me showed me some of the ways from which the site is able to recommend me thing I like.

    \\begin{itemize}

    \\item[1]- Purchased shopping carts = real money from real people spent on real items = powerful  data and a lot of it.

    \\item[2]- Items added to carts but abandoned .  

    \\item[3]- Pricing experiments online where they offer the same products at different prices and see the results.

    \\item[4]- Wishlists - what$'$s on them specifically for you - and in aggregate it can be treated similarly to another stream of basket analysis data.

    \\item[5]- Referral sites (identification of where you came in from can hint other items of interest).

    \\item[6]- Ratings by you or those in your social network/buying circles - if you rate things you like you get more of what you like and if you confirm with the "i already own it" button they create a very complete profile of you.

    \\item[7]- Demographic information (your shipping address, etc.) - they know what is popular in your general area for your kids, yourself, your spouse, etc.

    \\item[8]- User segmentation = did Ii buy 3 books in separate months for a toddler...and i got a recommendation for the fourth one.

    \\item[9]- Direct marketing click through data - did you get an email from them and click through? The website knows which email it was and what you clicked through on and whethever you bought it as a result.

    \\item[10]- Number of times viewed an item before final purchase

    \\item[11]- Your history on the website.

    

    \\end{itemize}

   \\begin{center}

    \\includegraphics[width=2.5in]{img/10.jpg}

   \\end{center}

\\section{Conclusion}

\\lettrine[nindent=0em,lines=3]{R}ecommendation systems are a very powerful technology for extracting additional value of data from its user databases. These systems help users find products they want to buy or see from a website. \\\\ \\\\ Recommendation systems benefit users by enabling them to find products they like. Conversely, they help the business by generating more sales. Recommendation systems are rapidly becoming a crucial tool in E-commerce on the Web. \\\\ \\\\ Recommendation systems are being stressed by the huge volume of user data in existing databases, and will be more stressed by the increasing volume of user data available on the Web. New technologies are needed that can dramatically improve the scalability of recommendation systems because in my oppinion those systems will become old-fashioned over time and wouldn$'$t handle a bigger data wave

...(download the rest of the essay above)

About this essay:

This essay was submitted to us by a student in order to help you with your studies.

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay Sauce, . Available from:< https://www.essaysauce.com/essays/marketing/2016-5-26-1464223198.php > [Accessed 21.10.19].