by. Sowa Asst: Pdijessov eS Pst of Techn SREYAS INSTITUTE OF ENGINEERING AND TECHNOLOGY UNIT 1 NOTES ‘TA SCIENCE APPLICATIONS Syllabus: Data Science Applications in various domains, Challenges and Opportunities, tools for data scientists, Recommender systems — Introduction, methods, application, challenges. Introduction to Data science : Whatis data science? Data science combines math and statistics, specialized programming, advanced analytics, artificial intelligence (Al), and machine learning with specific subject matter expertise to uncover actionable insights hidden in an organization’s data. These insights can be used to guide decision making and strategic planning. The accelerating volume of data sources, and subsequently data, has made data science is one of the fastest growing fields across every industry. Organizations are increasingly reliant on them to interpret data and provide actionable recommendations to improve business outcomes The data science lifecycle involves various roles, tools, and processes, which enables analysts to glean actionable insights. denial a data science project undergoes the following stages: Cancel a alyae ec) Data AS fon: The lifetycle begins with the data collection--both raw structured and unstructured data from all relevant sources using a variety of methods. These methods can include manual entry, web scraping, and real-time streaming data from systems and devices. Data sources can include structured data, such as customer data, along with unstructured data like log files, video, audio, pictures, the Internet of Things (IoT), social media, and more. Data storage and data processing: Since data can have different formats and structures, companies need to consider different storage systems based on the type of data that needs to be captured. Data management teams help to set standards around data storage and structure, which facilitate workflows around analytics, machine learning and deep leaming oOdeduplicating, transforming 1m, load) jobs or other tion is essential for models, This stage includes cleaning data, and combining the data using ETL (extract, transfor data integration technologies: This data preparat promoting data quality before loading into a data warehouse, data lake, or other repository. Data analysis: * Here, data scientists conduct an exploratory data analysis to examine biases, patterns, ranges, and distributions of values within the data, Far al tin «This data analytics exploration drives hypotiedis generation fora fot owlity adXurcincetesting, It also allows analysts {0 determine the data’s relevance for sar ise within modeling efforts for predictive analytics, machine ah tery * learning, and/or deep learning. aaa ¢ Depending on a model's accuracy, organizations can become reliant on these insights for business decision making, allowing them to drive more scalability. 3¢/- Communicate: wsineey Finally, insights are presented as reports and other data visualizations that make the insights—and their impact on business—easier for business analysts and other decision-makers to understand. eA data science programming language such as R or Python includes components for generating visualizations; alternately, data scientists can use dedicated visualization tools. Data science versus data scientist Data science is considered a discipline, while data scientists are the practitioners ‘within that field, Data scientists are not necessarily directly responsible for all the processes involved in the data science lifecycle. typically handled by data engineers—but the For example, data pipelines are data scientist may make recommendations about what sort of data is useful or required. While data scientists can build machine learning models, scaling these efforts at a larger level requires more software engineering skills to optimize a program {0 run more quickly. AS a resull, it’s common for a data scientist to partner with ‘machine leaming engineers to scale machine leaming models.extract insights from big data using pred ing intelligence (AI), including machine learning models, natur fF processing, and deep leaning. ||—and illustrate that clearly convey th t jon-makers and stakeholders at every I f nica Explain how the results can | to solve business problems business analysts, IT architects, data eng accion! ing into a data science career, explore a variety of dat ve programs 2 versus busine: 1telligence(Wp Tools: MS txcel » rods: Python , R Aladoop, “ertordia Y fase science deals sith rsh de on gl ‘ Predic tue ortllys!s (tara Could apres} (already ha ' to confuse the terms “data science” and ‘business intelligence > Itmay be easy and analysis of (Bl) because they both relate to an organization's data data, but they do differ in focus. Business ileligence (8) is typically an umbrella term for the technology that ee ‘data_mining, data_management, and data Business intelligence tools and processes allow ‘end users to identify ‘actionable information from raw data, facilitating data-driven decision-making within organizations across various industries ‘While data science tools overlap in much of this regard, business intelligence focuses more on data from the past, and the insights from Bl tools are more descriptive in nature, It uses data to understand what happened before to inform a course of action. Bl is geared toward static (unchanging) data that is usually structured. While data science uses descriptive data, it typically utlizes it to determine predictive variables, which are then used to categorize data or to make forecasts Data science and 8! are not Jusive—digitally savvy organizations use both to fully understand and extract value from their data. Data science tools: Data scientists rely dn popular programming languages to conduct exploratory data analysis and statistical regression, These open source tools support pre-built statistical modeling, machine learning, and graphics capabilities, +R Studio: An open source programming language and environment for developing statistical computing and graphics. + Python: Itis a dynamic and flexible programming language. The Python includes numerous libraries, such as NumPy, Pandas, Matplotlib, for analyzing data quickly. Data science and cloud computing Cloud computing scales data science by providing access to additional processing power, storage, and other tools required for data science projects. Since data science frequently leverages large data sets, tools that can scale with the size of the data is incredibly important, particularly for time-sensitive projects. Cloud storage solutions, such as data lakes, provide access to storage infrastructure, which are capable of ingesting and processing large volumes of ‘These storage systems provide flexibility to end users, allowing them to spin up large clusters as needed.a 1Iculate and p sing me tical 1 a you already kr fodeling helps in determinir i sIgorithn andi tain issu W r B fits of o in Busin \pplications of Data Scienc: Y r nendatic hnique can influence customers tobundled shornpes and canal “Furthermore, customers wll buy them Together fer a a ! Forecasting itis. one of the widely ‘applied techniques in Data Sclenes on the basis Votious types of data thot are caliseted fram varloun sources weather forecasting and future foracasting are done. Fraud and Risk Detection itis one of the moat logical applications of Data Science, Since aniline ‘transactions are booming, losing your data Is pessitla, For example, Credit card fraud detection depends on the amaunt merchant, location, time, and other variables, any of Ihem Fook Unnatural the transection will be cutormatically canceled, ard it will Blok your card for 24 hours or mare. selt-priving Cor The eali-driving car ig one ot the most euccessful Inventions 4 todays world. we train our car to make dacisions Independenty based on me pravious data, in this process, we com pence ait ‘model if it does. not erform well The car becomes more intelligent ‘with time whan it start learning through ail the reai-lime wxperiencas, Image Recognition When you want to recognie same Images data science can detect the object and ctoselfy it The most formous example of image recognition 1s face recagnition — If you tell your gmariphone to unblock it it will sean your fare. So lirst, the system will datect the foce, thon classify your face a3 0 human face, and after that, it will dacide if the phon belongs to the actual owner oF net. ‘ ‘speech ta text Convert “speach recognition js 0 process of understanding natural language by the computer, We are quite tarnifiar with viru! assistants like Sir Atexc, ancl Google Assistant2. De ical Once you've gained the foundatists Understanding of data science, you'll need to develop practica) skills that will come in handy in your career. For instance, familiarize yourself with programming languages, like Rand Python, and coding and database management systems. You may also want to practice machine learning and data analysis techniques. 3, Earn a Post Graduate Certificate or a Degree: Most employers prefer to hire data scientists with a post-graduate or master's degree in a corresponding field, like computer science or applied mathematics. Earning a Data Science or Analytics degree can help you acquire the knowledge, expertise, and skills required to become a successful Data Scientist. liands—on 4. Work on Projects: One of the best ways to develop your Data Science is to work on projects. You can find projects online or Teach out to organizations looking for Data Scientists. Working on projects will help you gain experience in data analysis, machine learning, and other Data Science activities. 5. Stay Up-to-Date: To stay ahead of the curve, you'll need to stay in the know about the latest Data Science trends. Keep an eye on industry news and subscribe to prominent pata Science Publications Data Science Applications in various domains » Data science is a new area of research that is related to huge data and involves concepts like collecting, _preparin, visualizing, managing, and preserving. > Even though the term data science looks related to subject areas like computer science and databases, it also requires other skills, including non-mathematical ones. » Data science not only combines data analysis, statistics, and other methods, but it also includes the corresponding results. » Data science is intended to analyze and understand the original phenomenon related to the data by revealing the hidden features of complex social, human, and natural phenomena related to data from another point of view other than traditional methods.Data science includes three stages: 1.Designing the data 2.collecting the data 3.finally analysing the data. There is an exponential increase in the applicability of data science in various areas because data science has been making enormous strides in data processing and use. Business analytics, social media, data mining, and other disciplines have benefited due to the advance in data science and have shown good results in the literature. Data science has made remarkable advancements in the fields of ensemble machine learning, hybrid machine learning, and Sees Machine learning methods (ML) can Jearn from the data with minimum human interference. Deep learning (DL) is a subset of ML that is applicable in different areas, like self-driving cars, earthquake predictions, and so on. There are many pieces ‘Of evidence in the Titerature that show the superiority of DL over ML methods; DL methods include artificial neural networks, K-nearest neighbors, and support vector machine (SYM) in different disciplines, such as medical, social media, and so on Ser Data Science Technological tools developed recently over the years have helped in many domains, including management and big data. Advancements in( || known that data scie Juilds algorit! nd syster i r patterns, ani ating useful inform do so, it encompasses an entire data analysis proc t d cleanin tends to nalysis, descript ind ‘ of data and cleaning d t 5, vith da ior Ir iz any the data is prepared, an exploratory analysis that includes visualizing tools will help decide the algorithms that are suitable to gain ‘i the required knowledge. This complete process will guide the user toward the results that will help them make suitable decisions, Depending on the primary outcomes, the complete process should be fine tuned to obtain improved results. This will involve changing the parameter values or making changes to the datasets. These kinds of decisions are not Thade automatically, so the involvement of an expert in result analysis is @ crucial factor. From a technical point of view, data science consists of a set of tools and techniques that deals with various goals corresponding to ‘multiple situations. ‘Some of the recent methods used are o ot > Clustering > Classification > deep learning > regression > association rule mining > time-series analysis. Even though these methods are often used in text mining and other ®t anomaly detection and sequence analysis are also helpful to provide excellent results for text mining problems. ‘(classification we have classified a set of objects that predict the classes based on the attributes. Decision trees (DT) are used to perform and visualize that Classification . DTs may be generated using variougsafgarithins, such as 1D3, , CA,.5, and C5.0. ‘Tkvatve Ofc! caesar Ganbeiory Sse pan Ker eae IRE is one more classifier that will construct a $#% of DTs, and then predicts through the aggregation of the values generated from gach DT. A classification model was developed by using @ technique known as Least Squares Support Vector Machine (LS-SVM).The classification task is performed by LS-SVM by using a hyper. multidimensional space for separating the dataset into the target class ‘WpRegression: Statbbcal method shows velabionshiP bet” tus snore Vanlables. 09: gerbe: Reson hit » agen Regression analysis aims for the numerical estimation of the relationship between variables. This involves the estimation of whether or not the variables are independent. Ifa variable is not independent, then the first step is to determine the type of dependence. Chatterjee et al. proposed a regression analysis that is often used for Predicting and forecasting, and also to understand how the dependent variables will change corresponding to the fixed values of independent variables . 11))Deep learning Deep learning is a branch of machine learning which is based on artificial neural networks. It is capable of learning complex patterns and relationships within data. In deep learning, we don't need to explicitly Program everything, It has become increasingly popular in recent years due to the advances in processing power and the availability of large datasets Because it is based on artificial neural networks (ANNs) also known as deep neural networks (DNNs). These neural networks are inspired by the structure and function of the human brain's biological neurons, and they are designed to learn from large amounts of data. 1. Deep Learning is a subfield of Machine Learning that involves the use of neural networks to model and solve complex problems. Neural networks are modeled after the structure and function of the human brain and consist of layers of interconnected nodes that process and transform data 2. The key characteristic of Deep Learning is the use of deep neural networks, which have multiple layers of interconnected nodes. These networks can learn complex representations of data by discovering hierarchical patterns and features in the data. DeepsadAy e - sasse} 3*! Se uy aue|t- Learning algorithms can automatically learn and improve from data without the need for manual feature engineering, 3. Deep Learning has achieved significant success in various fields, including image recognition, natural language processing, speech recognition, and recommendation systems. Some of the popular Deep Learning architectures include Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Deep Belief Networks (DBNs). 4, Training deep neural networks typically requires a large amount of data and computational resources. However, the availability of cloud computing. and the development of specialized hardware, such as Graphies Processing Units (GPUs), has made it easier to train deep neural networks. piss Clustering proposed a clustering-based method using the degree of similarity - In clustering, the objects are separated into groups called clusters. This type of learning is called unsupervised learning, as there is no Prior idea cover the classes as to which group the objects belong. Based on the similarity measure criterion, cluster analysis has various models: (i) based on the connectivity distance, connectivity models are generated, i.e., hierarchical clustering; (i) by using the nearest cluster centre, the objects are assigned, centroid models are generated, i.e., k-means; (ili) by means of statistical distributions, the distributed models are generated, i.e., expectation-maximization algorithm; (iv) based on highdensity areas that exist in the data, the clusters are defined in density models; (v) graphs are used for expressing the dataset in graph-based models. » Association rulesi aup syeoiput + ayy oye OV} oy qeip vont ‘Clamifieation Teetoiqus (OMB. RY, DT tal) spagenion Tecinique * Antiicial Neal Network Techmiquee FIGURE 12 Data Science Techniques Applications of Data Science in Various Domains Data science is one subject that has gained popularity out of necessity, corresponding to real-world applications as a substitute to research domain Its application began from a narrow field of analytics and statistics and has improved to be applied to different areas of industry and science. Consequently, this section explains the data science applications that can do the following: (i) economic analysis of electric consumption, (ii) stock market prediction, (iii) bioinformatics, (iv) social media analytics, (9) email mining, (vi) big data analysis, and (vii) SMS Mining, among other thing Economic analysis of electric consumption: Economic Analysis of Electric Consumption Different electric companies or utilities approached data science to find out and understand when and how consumers use energy. ‘There has been an increase in competition among companies that use data science to develop such information.‘Traditionally, this information has heen determined via classification, ‘and_patiom analysis methods by using the association rule, have ‘consumers As various classes based on their behavior and usage of electricity, The comparative evaluation was made with sel-organizng maps and an improved version of follow-ihe-leader methods, This was the fist step initiated for a tariff of the electrical utilities. Figueira have develaped a framework for exploiting the historical data, which consists of two modules: (i) 2 lond-profile module, which creates a set of customer classes by using unsupervised and supervised learning, and (il) @ classification module, which butlds models for assigning customers to their respective classes, Stock Market Prediction An application of ML and DL techniques in the stock market Is Increasing tompared to other areas of economics Even though investing in the stack market gives profits, high risk is often involved along with high benefits, So, Investors try ta estimate and determine the value of stack before they make an Investment ‘The cost of the stack varies depending upon factors like local politics and écoriomy, which causes alfficulties im. identifying Future trends of tRe-Stock market, a Fischer and Krauss sed [STM to forecast future trends in the stock market. The results have been compared with LOG, DWN, and RF, and have shown improved Fesults ower the others. Tamura et al. have proposed a new method for Predicting the values of the stock, Here, financial data related to the stock market of Japan has been used as a Prediction input in LSTMs (Long short-term memories). Further, the financial statements of the companies are recovered and then added to the database. Sharaf and Srinivasarao proposed Linear Support Vector Machine {LsvM) Identify the Setrelation among the words in content and subject of the emailsBig Data Analysis! Big data is one oft the present er@- The instrumental variable regression technique has been used to analyze Facebook data. Here, the emotions of the people, such as negative and positive emotions during rainy days, were detected. Roelens et al. They explained that the detection of the people who influence social networks is a difficult task or area of research, but one of great interest so that referral marketing and spreading information regarding products can react the maximum possible network. E-mail Mining aCThere is a threat to internet security with spam emails. Spam emails are nothing but unwanted or unsolicited emails. Mailboxes will overload with these unwanted emails, and there may be losses in storage and bandwidth, which favors quick, wrong information and malicious data. Gudkova et al. They conducted a study and explained that 56% of all emails are spam emails. Caruana and Li illustrated that the machine learning method is successful for detecting spam data) %e These include learning classifier models, which map data by using features like n-gram and others into spam or ham classes Dada et al. have demonstrated that email features may be either manual or automatic, Bhowmick and Hazarika demonstrated that the manually extracted rules are known as knowledge engineering, which requires expert and regular updates to maintain good accuracy Text mining methods are used for automated feature extraction of useful information like words, enabling spam discrimination, HTML mark up, and so on. Using these features, an email is represented as Bag-of-Words (BoW) as proposed by Aggarwal Here the unstructured word tokens are used to discriminate the spam messages with the others. The BoW assumes word tokens that are not dependent that will prevent from delivering the good semantic content to represent the email. Sharaff and Nagwani have identified the email threads using LDA- and NMF- based methodology.‘| SUR Big Data Analysis Mining Methods Big data is one of the very fast-growing technologies that is critical to handle in the present era. The information is used for analytical studies to help drive decisions for giving F quick and improved services. Laney proposed that big data consists of three characteristics: velocity, volume, and variety>) These are also called the 3Vs. They explained that data mining is a procedure where potentially useful, unknown, and hidden meaningful information is extracted from noisy, random, incomplete, and fuzzy data. The knowledge and information that has been extracted is used to derive new comprehensions, scientific events, and influences business scientific discovery. Two articles have aimed at improving the accuracy of data mining, It have proposed a new model using the skyline algorithm. Here, a sorted positional index list (SSPL), which has low space overhead, has been used to reduce the input or output cost It shows an overview of data science methods used in different applications.TABLE 1.1 ; : ‘An overview of data science methods usd in different applications ceed ipicaies Methods ce 1 Economie Follow-the-Leader Clustering (FLC) ico e a Ries Figueiredo etal. [15] (1 Fischer and Krauss (6) 2 Stock Market Long Short-Term Memory (LSM) Tanna dal) ‘1 ing (GDL) Baldiet al. (17] support vector machine (SVM) ‘Ambroise et a. (19) 4 Social Media Naive Bayes (NB) and Maximum Entropy Toshi Sena fat) analytics rithms (MEA) ay eee ear Coviello et al. (23) Regression Methods (RM) : 5 Email Mining ‘Machine and Non-Machine Learning ‘Caruana and Li [26] Methods (NMLM) Dada etal. 27] Deep Leaning Methods (DLM) ‘Bhowmick and Hazarika Machine Learning Techniques (MLT) 23) Latent Dirichlet Allocation and ‘Sharaff and Nagwani (30) ‘Non-Negative Matrix Factorization (NNME) 6 Big Data Fuzzy Clustering (FC) Chen et al. [32] Analysis Data Mining Methods (DMM) Liu [33] Skyline Algorithm ($A) Han etal. 34] |-4 Challenges and Opportunities \-4-\ Challenges in Mathematical and Statistical Foundations The main challenge in mathematical fields is to find out why theoretical foundations are not enough to solve complex problems, and then identify and obtain a helpful action plan. \-4-2 Challenges in Social Issues In social contexts, the challenges are to specify, respect, and identify social issues. Any domain-specific data is to be selected, and then its related concepts—like business, security, protection privacy—should be accurately handled. \-4/3 Data-to-Decision and ActionsAM y : } k is i i a mportant i develop accurate decision-making systems fe are data-driven. These systems should also be able to fanage and govern the decision-making systems. |-4-4Data Quality Enhancement Another important challenge is issues of data quality like uncertainty, noise, unbalance, and so on. The level of presence of these issues will vary depending upon the data complexity. |g. Deep Analytics and Discovery Cao [35] proposed new algorithms to deal with the deep and implicit analytics that are not able to be tackled using the existing descriptive, latent, and predictive learning. Also, how to aggregate the model based with data-driven problem- solving solutions to balance the domain-specific data complexity, intelligence-driven evidence learning, and common learning frameworks. bqe6 High-Performance Processing and Analytics Systems must handle the online, real-time, Internet-based, large-scale, high-frequency, data analytics and processing with balanced resource involvement that may be local and global. This requires new array disk storage, batch, and high performance parallel processing. It is also necessary to use complex matrix calculations, data-to-knowledge management, mixed data structures, and management systems. High-Performance Processing and Analytics} x ©Systems must handle the online, real-time, Internet-b; large-scale, high-frequency, data analytics and process with balanced resource involvement that may be local and global. This requires new array disk storage, batch, and high performance parallel processing. It is also necessary to use x complex matrix calculations, data-to-knowledge Management, mixed data structures, and management systems. Cay Networking, Communication, and Interoperation The challenge involved is how to support the interoperation, communication, and networking between various data science roles like distributed and complete cycle of problem- solving in data science. Here, it is necessary to coordinate Management of tasks, data, workflows, control, task scheduling, and governance. }-5Tools for Data Scientists It presents the tools required for data scientists to address the aspecis discussed above, These tools are classified as data and application integration, cloud infrastructure, programming, visualization, high-performance processing, analytics, master data management, business intelligence Teporting, data Preparation and processing, and project management. The researcher can use any number of tools depending upon the complexity of the problem being solved, '-S\Cloud Infrastructure Like Map R, Google Cloud Platform, Amazon Web Services, Cloudera, Spark, Apache Hadoop, and other systems may be used. Most of the traditional IT vendors at present are using cloud platform.Cloud infrastructure refers to the hardware and software components, such as servers, storage, networking, virtualization software, services and management tools, that support the computing requirements of a cloud computing model. Cloud infrastructure also includes an abstraction layer that virtualizes and logically resents resources and services to users through application programming interfaces and API-enabled command-line or graphical interfaces. Servers Major public cloud providers, such as Amazon Web Services (AWS), Microsoft Azure and Google Cloud Platform, offer services based on shared, multi: tenant servers. This model requires massive compute capacity to handle unpredictable changes in user demand and to optimally balance demand across fewer servers. As a result, cloud infrastructure typically consists of high-density systems with shared power; often, these are multisocket and multicore servers, (5-9 Data/Application Integration This includes Clover ETL, Information Builders, DM Express Syne sort, Oracle Data Integrator, Informatics, Including Ab Initio, and so on. \ 5-3, Master Data Management Master data monagement includes SAP Net Weaver Master Data Management tool, Black Watch Data, Microsoft Master Data Services, Informatica MDM, TIBCO MDM, Teradata Warehousing, and soon { 5:4. Data Preparation and Processing stodder and Matters have used some platforms and data preparation tok ke Wrangler Enterprise and Wrangler Alpine Chorus, BM SPSS, Teradata Loom, Patfor, and so on 1-55 Analytics “Analytics includes commercial tools like Rapid Miner [37], Mat Lab, IBM SPSS Modeler and SPSS Statistics, SAS Enterprise Miner, and so on, in addition to some new tools, like Google Cloud Prediction ‘API, ML Base, Big ML [38], Data Robot, and others.|-S-¢ Visualization "1 Some commercial and free software listed in KDnuggets 139] to visy rs include Miner3D, IRIS Explorer, Interactive Data Language, Quadrigram, Science GL, ana so on. | -5-F-Programming Additionally, Java, Python, SQL, SAS, and R languages have been used for data analytics, Some data scientists haye also included Go, Ruby, .net, and Java Script . ) S$ High-Performance Processing Around 40 computer cluster software programs, like Platform Cluster Manager, Moab Cluster Suite, Stacki, and others, have been listed in Wi ipedia (41). 5-4 Business Intelligence Reporting Some of the reporting tools commonly used are SAP Crystal Reports, SAS Business Intelligence, Micro Strategy, and |8M Cognos, among others \5-JoSocial Network Analysis Around 30 tools have been listed for social network analysis and to help visualize data, For example, Ego Net, Cuttlefish, Commetrix, Keynetiq, Node XL, and so on. Figure 1.3 shows the different types of programming languages that are used in data science. oe B8SS8259S ef yh Pap ne | Be =a, on bq data fas been tx > Data Science wicthns abl) be extensively eed ena Pelsure Qita soence echniawes voveal, eupoant tools that ean extract | exploit inte mmabon y Knocoledpe rae ers a ome dataset - aRecommenler Systems Bek Qotrodyetton: Read. by user iv Simlor ay Hcles Seg Recommended. +p 0Seu: fH vecommenden. Sapiens ane highly useng discover Procects € vse tol a3 they het P lhewe. not ound Seyi 6 & they mig he othere on their OWN RS Ot trainedto Understand. +e. Pre seventers ? Previov cluvacleskes 06 people. and Prod ucts USIOF dato Greve! abot Prete Vortrac ons: : 5 decisions ane |24. aybid fy 2@> Yodren Recommend. Sqtems. 4 e- Ddla — Driven Recommendations. 5A. Sate knowledge ~Dyiven « 4 6B Cogrition— Driven Recommenda Ho ns. Q.2\ epee: Collaborate ering : SDL is Used im almost all application domains. > Thy method docused on ebbective adopPtten 06 the, sen deed back (ey: Ratings) $rorn oser3 40 emake. a probile ob advimities generate Personalized = Sucn Profiles one osed to rpecormmenda Hons: Predicated vatings » Collaborative. fi tecing igems based on Predicted Aa ngs cthese with the highest ya tings. 4 Haut can sort sthe and arecornmend ods in collaborate Mg Systems — bared » which comp~te vier to-ose m_ Similavibes heteversy bared on faterns ob the Uos and itens- > ctogtcal meth ane. neigh boo or Wem tot dhe. con va tig > Fence. » Ta begs hew the. ‘tem hag been who Were considered (ike -minded COmRued ty the tonaet User: Z predichons 2s Pertorrred bated on co-voted by other Useng& Pnosthey — Syvtenns adopl bab c order sto genererle, a 9 > A well, adopted carlegory © Anevtix dactoriga ton, nae oy Vlaiix dactorigation. decomposes the raticg Yodo dy btevent abeie ese nee ‘i @ Whave 5 tSmahix ob TAKE 5 M is matrix ob JC) XE ® Where. Sug Pescri bes level of User LA Prekvenc +lowards the factor F » desevi beg Strength of the factor + baale Asia che ‘ttt on a QQ co. tent = Based Recom me ton : rau an > Conte d methods % loPt conttot — base ondent Base e do? Mt Cert ng algorinms tn ov prelenee | Cees profiles Fy axed hy Tem. cowtent: > contar based methods waagure a velevaney Score. agocated Ho OSC Pre ferences Propordonal. 10 the Content Jeadres. :— & Accordicg to the. Pre serences , Als tikes ov Sta Ser the. idems Neigh boor > set cHye Pye #re nce ani chen Used 70 A, bor user % and X15 959))7 New ese we ny = ae Ejewni sayz Sil? 0 vetects The elonmensts db~ where Yu 7 Le User arabes Incloded. Preference, matrix R> In the ‘Wasdix ob all rakngs | “e J aad pe so + Pn order 70 addras caetictons or Colla bovati« ‘ ert ag emothnod and correo t — baxed method » Barina en athods been developed by evesen nae ds: n algo have, aSvene rmac hi nes: have ya rmesthods CA factoiaaton ee prekerenc es (a, like g ered ce athe sey 7 a 2) condecttal sthyoog hh FNL formok driers oF 228 Pee denotes bie Yactor we CS user wet ancy DBPLIE of ebbecttvency ob nd AD the dy 13s been © Strong = torre ob ecommen der eng eS: > 0) acniaue) me gook > smo dren Kons Stems hove. (ees) 3 moth specks oe Dake eK nowledg @ mi Gon: no Command: abon oN ray of) DOW ives tslands ? | soctel qbba source | poe ore qenecatd open] ane | yoclod re . ig oprawchs Wen core. DB ProPse 4 potion ob dete lala sere am ans 4 aod oyue. 4A \osqe Beant ob Angocmatter. ( brn fer» dq Private data) ond to fer) Soy o8 to deal worth, the vanteby enable. analysts be data. ond ona stancdad data, medals.Rab knowledae Haven Reco kenowledge, taker Ck Us) +wo challenges : ToeleqenE te olloes ’ ty the. Col ~ start blems leveraglora 4 oell cs 4 tm bor mabive knowledge. toke, wi . daka berm eo ok People and Vie ie | to Generate recommen ae gies ood Variance ® revevaging Birtellegiend= knowledge Janes be able, to qoide the wecommmender 245 bask next steps (ear red ons: ual l | tems tO choose, by filo 9g bert brome domeim 2¢€ Pets: Pract (24 apis can be done, by- osing tee hntawves Sul 4 collecting ted back Organiaing Deterviews- : mpeoyues BG ads: =o achteve T's gook rib tS Tm Poder Peeicceae means exerts and en Hes Gnas ships among thom) | | hat and happen Rin ei ane venoes axcielinc> aye wate | édecaton fad bin teh. $8 ech agfet Coa nbon= Deven gy = OY ts vital Yor a yecommendey System Slemilax te, : , ae 4 bared on, theory be haviour, e& and Cao\ ve Ho} oking . f Mrtellegence f dniellectual. > factittate understanding users pexsorall ten 5 ermotfons 9 moods and > «Himitie, oven Gime US this Lage alans to emrPown. He. Secommendr— todels £4 explottation ob “Cognitive signals @nd meuval data: WS A Cogni tve Rs may docug Qn. dimensions Such ag expliak behavioral, Pattern. § tan plete behavioral, Patterns: | ; ees WP caper Pavers med melede Take age tocation,— based approaches » aclon— based toate : fa Nee a | and deature. - based methods. B23 Applicators: us Multimedia Ue ao ne classic wp Food ivy Fashion > i> Fieancial Jechnology CHtntech) ih & ducation Mod ven i> Recyuf -menta rs Sap on i | iy Mutttamedias Multimedia yvecommenderS ; 4 exploit dibtevent — fovrns ok Wy and Can Use. dibtevent- a, descxtptors wohen Stems ean Pretevence, data xy Pos ob muttimed? Sa xe comrendatons + “> such features ane classibied tmto 2 mein Cacteapovies e oF avigh= level torm is Low ~Level torm tevel deceriPtors ‘llushate more ob peer 30d eyntalle Charactersti cg ob amubtiomedic reteetand: C20 be Aq atd fom either s ob ‘meta data: e aah Ahuctored forrn. Melecols descrifftors ONC. Sq qty tmedia, piles (eq: audio or visual tiles) tea divectl dom mut +h # Low level. desea Ptors Can swpresent the acoustic i os (€: het hen: 5 energy § melody Con bigura-Hons recommender foliel can) CS adopted syctems to bend Sinmila Songs and to cree Personalized ve commendations tor user.EBS erodes _ atextuali gation, é Investigate oe fo an oe th : pe) Contextual, factors like Weather Condi tons Foravel Goals . ; * vied ob Transport ation- Ye the Secommendadtor qenera ton ~The idea ‘s-+to make Persoral sugaestions a imCor porating diverse sources ob User cta as well as che Condition ‘ePpresented ae Contextual, sactors: > for eq A grove ol tourists nae 4 in visi Su gestions to doov athactions back Weathex bz nice pPretew outdoor activites be interested | Ces ynusem S) during weather. tne ie | (ey. ntxta) S > Recommen der System Such Contextual factors Ave (Conxxt Aware Recommender systems] ‘ Rs ave empowered to explore emarthe malice! a peace] romance learn user Roraentt m™ 4 tm didtevent Contixteal, Sttva 4 ons z . that ave capable ob ve known a8 CARS) ' etevences fe a data, Peres om diverse, Sovrtes 0Tem pera tere %® Season ¥ Geog va pri cal. Positien— * 9 even. velicle tyfe- FMW Pood: the, Pretevencen that art. aggregeuc ee ee Corn mendey spum eee Whew, Cane (ame xe Long -deem afd ni Hes ¢ Q Ueking— se Short —term. abbini tes: sUhile obtainin and Baie pissin Caras eh Prebcvencer, 2 eential- ? ‘ | > the yeroorth on re common.dev Sqsterns dhol Identity the. didtevences between. thee two tore, teed cation? > On & te baat ‘sve. Is growing dives ty y ex Panding number ob fashion. Paodocts- o> nis & an. ebfect that Can Certainly (5A0 5 Alea Deen Pend cx aL eaeBy mock me cerésar' ce More quaila ble oe Puhey iikelihcod | aheemrenees md a deaived predcae ; uch. an ebkect mi Leeda loratees tem Posty of actually “real a Prodeck Pe the Problem ob receiv foo rang OPHONS » _Particalendy w howe a ann, we ‘diverse - echnioves ane Powey bol tools that can estective _ tackle this 7Ssue by maki ng yelevant Sysqestiors ob Products Lailored to the weeds the, USers- HE sthey can buritd a sitter ae a ! i evay i ea eA on nate, unt ncteres +t an yee (eu: => Recommender ’ Ms recommendatio I short list Products from % r ecsrene SA behind the i ence > iL cam, acted Understand the mend ey di el nme com Roxie and Obteti on. ere an : Users : ee ot cae descyi bing vse 4eq tor ma et eye Bie obi. knowledge on basta, im OY We he. aad deal doers OF ay8 Banking en Besssvchas customers g Ee 5 Ranking dornato knowl @ Gvchas how cite = mt banking S ¢ nk Operade ACross Customers 9 Sales and distriboton , Products % “Servs cey, i aye > Bank\ Processes, +0 hel P and, understand tre best eyes learned oe Knevolede experts IM Processeag Such as ' dreox Freawd detection oa Customer segrrentabon People. » Processes .And 1 h degunet a castorney Ce ba RE mode li ovesttran b baldes: ete ~ - a porn short comt oF exis ny -e5's i hak de ah dont consider, domain cxPerts knowledge By Ne a OL not exploit Uer-side into avacter sty of “the sen —Several ee an education ¥ Prsteuctors *¥ Workload, én dealing sith asessments And § Rovidi recommendations bused 90 Stodants. Pertormance, and Skills Assessmmne + wil in sthis Condext» Yecormmender Ssierns can be Geni cantly Imfoytank tools tor Personal iain: teaching, & canbe be ovderstandt and. Araly ging. "een Portank ‘indicators ae Skills. Cote ; Per toy mance, and . | d lanned work tor sthe, tutvre Could f | . aware, deeP learnt templemant a Lame, ue lear nes, ; constwct Y esata i | ae ia poeo belo. Understand Staents | Prosiles | : t Portor nance, and > An athactve P skills: | ‘ ’ wold enable ecommen 4, the learnto models epiems co “identity simf lar 4 Pec torr rg ato dont 7 which may facilitate Personalizing athe procers 7 Sobfect selections and recy wht ments + ~Kaations and Yecommend er Sprems Be . fon P gole, In aasisting yeeroiters in the ; vecruittent Process: &: Consider recommenda hon engine shat hos, access do Linkedin paotles, is able to extrack duke § Rectan. tom basiness arti bacts (C29: Candidates Cy g Position descriptions) Yas access to curation algo's +o cont xtual ze, Hu data, and — knowl ‘and 1s able to Ane thin to acts In wecyuitne Jomain Knowle i tment o Know daa > AL erabled orgenaa bors to Creaté busi ness, leverage, a ap egie asia edge, axtometen 4 sHehniqwues, Wy extracting levank information bron - Candidate cy autorrastically. a pers, Ms" 9 eee ment Proless- - E & clitheven.t 6a ndidat In bormetion iy othe yeow\E iy Aqqve 42 Hon. © ealuations § Te ‘Wy ondertendirg best Practices used by revues ~ dividual Chavactesiics | Li ait op a} Rider levaneSastem \ 1 that i Barlding e btec bee yankr hea: OP Hrnige ye Comrman dabons Henk: oes Patori by foo L of €x plo?ve $i Recommender Siptims dy Pical c ata Sets that Contain usec ae A. likey dis\ikes or Yatings) that : ~ePresent Preferences produced by & bly Croux O& Inter Connected users tO oO as lis Cot idems > explobn suc h data, CmMPowes Hye mecommender = “gat te +o learn tha Paternt | gq Penptacons wall users 9 and vse thew | ao eatimate the nissi asseaments O8 Osers toy unex plored yttros and stan S&S aye sdeng thak ™¢ be, attractive, to a farqee usey > ateve ave Sort Concerns. col stave : Batlle a ia weraied faye? ayoali me data: i > 4 sub- Problem of cold start fs Called the which veters to when & new at 1s strongly ear of tip sn eu USCY sitoati on »Sugqesnons Palor to giving « Preferences to 8rY existing System: eA now iter eftoation 5 a new fm is Natvodueed to the ‘em catteg. page Obtain axysasrent trom existing vsers mers Or 0G) 7 EU EDS Uy “> > Ttheve. exists GAnosthn, Problem Callod - ae - Qt 1S A Meagore @ data, oN and is ~Pofortioml to no ds aval lable teed backs Cea cary) over the overall Possible, teed backs. | pes FF of existing teed backs . lap we all Passible teebbacks! pO ob cold stave Condi Hors: Ae extreme. cold Start Ge moderate. oN naa AE Exteme cs: takes Place cohom User Starts Usin RIEU the. Stern and anes bor & Yecommendation hetore Prodocing any feod back . ssthe Problem can also apron when a bvand-ne = 72 Product Is inserted Into the @tloy. ang r Lee cSS Verto a. e lead +o tailove fn su qe edock tp an exist ba i, Ala wild) Hooerate cold start shaPeens once a Small no ob teed backs are. Produced pyar to Exist products g the Spt can oe, THis Wmited data, +o generat a sre commen dation: S This Problem “yay algo take Place for & now Produce when & Small amount ob content data, ave ‘0k Aly Produced - & may Geto, (Pale Cs > urid Codescay extreme, and wame Com bread condi 4en ot Sta 4e — This Can stil be Promptly addvessed by Vere aia system. B.g-d Comenk CEES: fieme Loca ben % iy Coual velabons ort examples aba Comext tor which am ofvated egeart hos TOY « Per enco) kadai dec h niayves suchagien | Leis peri . * ~ he sea ks oper oe 7. lead, +o a ta'lore, 1G nok| Sepe- 2 Ceometion of 24 ome). 4 weeks. ‘pelle, 23— 29> Oosheraveck 5S thetey | coy | thes teas totovial: ®@ Geta SS to 66pee ts Style. cwareness: 5 cc a domain 4 exam AS Ont ts Colorbolness § movement and | Sound - S thee, ane. diverse, Weasons bor havi. tahoe te Gling amd mmnouNes ee 4s to enable + - ob tHe Space & she porciued onderstand +o beild observa ble: oe thal audience ANE. ol Becca hn di ea, ose ae ele ne ocey yaree seen, ackny in 0 was arate 9o% Eaijood tog cal. pee eee to © force PROM ob & = pce 1) Com Panist otnev aes there Char ackyshs bone Honing 19 nna Nev - Ee co ae Ex Peres an nadia, have & Comm Reema tenfack oF colors becomes longer, as they is ing 2 fay dewlay emo! Bt——) shot dovation > i : ‘ % > Deapite. ob the VWmfortance, Ob lous level, deseriPtoys » tha. US ok thorn hax not craton. ynuch. Consideration 19 Ae commendatior Sy te ms +