Boutique hotels are unique and differentiated from traditional hotels, offering personalized service and a memorable experience to their guests. However, managing a boutique hotel can be challenging, especially when it comes to meeting the diverse needs and expectations of guests. A recent large-scale study of online reviews has shed light on the key factors that drive guest satisfaction in boutique hotels. Based on the findings of this study, here are 10 tips for managing boutique hotels:

1. Prioritize the Ease of Booking: The study found that ease of booking is a critical determinant of guest satisfaction, particularly in the US market. Ensure that your hotel’s booking system is user-friendly, the process is streamlined, and check-in is seamless. This includes optimizing your website for mobile devices, offering multiple payment options, and providing clear and concise information about your hotel and its amenities.
2. Focus on Service Recovery: While high-quality service is expected, the absence of expected service quality significantly decreases guest satisfaction(Mishra et al., 2024). Therefore, it’s crucial to focus on service recovery. Train your staff to handle complaints effectively, empower them to resolve issues quickly, and go the extra mile to turn negative experiences into positive ones.
3. Personalize the Guest Experience: Boutique hotels are known for their personalized service. Tailor your services to the individual needs and preferences of your guests. This could involve offering personalized recommendations for local attractions, providing customized amenities, or simply remembering their names and preferences.
4. Maintain High Standards of Cleanliness: Cleanliness is a basic expectation for any hotel, but it’s even more important for boutique hotels that promise a luxurious and comfortable stay. Ensure that your rooms and common areas are spotless, and pay attention to details like fresh linens, towels, and toiletries.
5. Invest in High-Quality Amenities: While boutique hotels may not offer as many amenities as larger hotels, the quality of the amenities they do offer is crucial. Invest in high-quality bedding, towels, toiletries, and other amenities that will enhance the guest experience.
   
   
6. Offer Unique and Memorable Experiences: Boutique hotels should strive to offer unique and memorable experiences that guests can’t find elsewhere. This could involve hosting special events, offering unique packages, or partnering with local businesses to provide exclusive experiences.
7. Leverage Technology: Technology can be a powerful tool for managing boutique hotels. Use property management systems to streamline operations, customer relationship management software to personalize guest interactions, and reputation management tools to monitor and respond to online reviews.
8. Monitor and Respond to Online Reviews: Online reviews are a valuable source of feedback for boutique hotels. The research paper by Mishra et al., 2024 provides step by step approach to collect and analyze guest reviews. Monitor reviews on platforms like TripAdvisor and respond promptly and professionally to both positive and negative feedback. Use negative reviews as an opportunity to learn and improve.
9. Engage with the Local Community: Boutique hotels should strive to be a part of the local community. Partner with local businesses, participate in community events, and support local charities. This will not only enhance your hotel’s reputation but also provide unique experiences for your guests.
10. Continuously Innovate and Improve: The boutique hotel industry is constantly evolving. To stay ahead of the competition, you need to continuously innovate and improve. This could involve introducing new services, upgrading your amenities, or simply finding new ways to surprise and delight your guests.

How to make the guest experience more personal and memorable?

Personalization in boutique hotels is a key differentiator that elevates the guest experience beyond simply providing a place to sleep. It’s about crafting a unique and memorable stay that caters to individual preferences, fosters a sense of connection, and ultimately drives customer loyalty.

The Importance of Personalization in Boutique Hotels

Boutique hotels, by their nature, offer a more intimate and curated experience compared to larger, standardized chain hotels. This smaller scale allows for a greater degree of attention to detail and flexibility in catering to individual needs.

Personalization is more than just remembering a guest’s name or preferred room type. It involves understanding their interests, anticipating their needs, and tailoring every aspect of their stay accordingly. This could include personalized welcome amenities, curated local experiences, bespoke dining options, or even customized room décor.

The Benefits of Personalization

1. Enhanced Guest Satisfaction: When guests feel seen, heard, and understood, their satisfaction levels naturally increase. Personalized touches make them feel valued and appreciated, leading to positive reviews and word-of-mouth recommendations.
2. Increased Loyalty and Repeat Business: A personalized experience creates an emotional connection with guests. They are more likely to remember their stay fondly and choose to return for future visits, driving repeat business for the hotel.
3. Differentiation from Competitors: In a crowded market, personalization helps boutique hotels stand out. It gives them a unique selling proposition that resonates with guests seeking a more tailored and memorable experience.
4. Higher Revenue Potential: Personalized experiences often come with a premium price tag. Guests are willing to pay more for a stay that caters to their specific needs and desires, leading to increased revenue for the hotel.
5. Data-Driven Insights: Personalization efforts often involve collecting guest data, such as preferences, interests, and feedback. This data can be analyzed to gain valuable insights into guest behavior, enabling hotels to further refine their personalization strategies and offerings.

Examples of Personalization in Boutique Hotels

• Pre-Arrival Questionnaire: Gathering information about guest preferences before their arrival allows for personalized welcome amenities, room setup, and activity suggestions.
• Handwritten Welcome Notes: A personal touch like a handwritten note from the hotel staff can make guests feel special and appreciated from the moment they check in.
• Curated Local Experiences: Offering bespoke itineraries and recommendations based on guest interests ensures they have a unique and memorable experience exploring the destination.
• Personalized Dining Options: Catering to dietary restrictions and preferences, or even offering private chef experiences, creates a truly bespoke dining experience.
• Customized In-Room Amenities: Providing personalized toiletries, snacks, or even a curated playlist can add an extra layer of luxury and comfort to the guest experience.

By following these tips, boutique hotels can enhance guest satisfaction, improve their online reputation, and ultimately drive more bookings. Remember, the key to success in the boutique hotel industry is to provide a unique, personalized, and memorable experience that guests will cherish and share with others.

This article is based on research paper Mishra et al. (2024) published in the Journal of Computer Information Systems and can be found here: https://doi.org/10.1080/08874417.2024.2362824

Mishra, A. N., Li, J., Sengupta, P., & Oztekin, A. (2024). NLP-Driven Insights on Boutique Hotel Satisfaction. Journal of Computer Information Systems, 1-16 https://doi.org/10.1080/08874417.2024.2362824

The post 10 Tips for Managing Boutique Hotels to Enhance Guest Satisfaction appeared first on Management Weekly.

It took almost two decades for e-commerce to gain the same number of users that m-commerce gained in just five years. There are some key differences between e-commerce and m-commerce that has lead to this phenomena. There are several differences but most importantly, we have portability, ease of use, integration with other services and ease of communication. These are some differentiating factors that set m-commerce on a different growth trajectory.

M-commerce stands for mobile-commerce. It is a derivative of the term electric-commerce of e-commerce. M-commerce can be defined as the exchange of goods, services and money among entities such as firms, customers, government supported by the use of wireless devices such as smartphones or tablets. Another defining factor is the use of wireless communication like Wi-Fi or Telecom internet. On the other hand e-commerce does not have the limitation of either the type of device or of the mode of communication.

Benefits of m-commerce over e-commerce

Mobility

There are many advantages of mobile devices. Firstly, mobile devices are always available with the users. You can simply take out your smartphone from the pocket and click on the Amazon icon. Then, you can start using their mcommerce application. However, if you want to use their website (ecommerce) then you may have to wait until you reach back home or office. You need to have access to a computer. Secondly, mobile devices more handy. Sometimes the ideas that you may get to purchase an item, may be transient. You could forget it by the time you reach back home and open your laptop. On the other hand, you can instantly browse the app.

One of the driving forces for the adoption of m-commerce has been the increasing number of mobile devices. Also researchers have found strong correlation between the factors that enhance or inhibit the mobile adoption to the factors for m-commerce adoption. What this means in simpler language is that mobility is the key to m-commerce. As we see a higher penetration of mobile devices, we shall see higher proportion of users going for m-commerce.

Ease of use

It is far easier to flick out your smart phone and start browsing items on Amazon. A key difference between e-commerce and m-commerce is that of ease of use. You don’t need to open the laptop, boot the system, open the browser then type the address of ecom website. You can simply click on the icon of your favorite ecommerce app and search for the item you need.

Also, the way, mobile apps are designed, there is a lot more standardization. It is easier to navigate another new app compared to another new website. The interface and buttons are designed to make your purchase journey as easy as possible. Although the same approach is taken on a website as well. However, websites are still loaded with sidebars, comprehensive menus, etc.

Integration with other apps

All the apps work over a common operating system. Whether it is Android which is the world’s largest mobile operating system or iOS, they all have mechanisms of connecting the data between different apps. This is much easier to do in mobile platform than desktop platforms. You can have specific files that open with specific apps. Similarly, the services can also be shared between different apps.

For instance, try to imagine a scenario where you need to return some defective item. If you are using ecommerce, then it is cumbersome to attach images of the product to your complaint. However, you can simply using the same (mobile) device to click images and upload it instantly.

Seamless payment

It is far more easier to integrate payment interface with a mobile app. The transition between shopping app to the payment interface is seamless. If your bank sends you an OTP, you can easily access that on your phone as you would be already holding it. Also, automation is simpler. Many of the apps have the option of reading your OTP. This makes the processes so smooth as you don’t even have to bother to read and enter the correct numbers.

Also, most mcommerce applications allow the local storage of your banking credentials. This can make the payment system faster. Finally, you may also get a confirmation message from the bank on your phone itself. You can check whether the payment went through properly from your bank’s end.

Ease of communication

M-Commerce enables faster communication between players. Firstly, you can use features such as notifications. They are also known as push notifications. These notifications help you in directly communicating with your customers. It is evident that they have much higher read rate, click rate and conversion rate compared to emails. Also, timely response from user is an added bonus. Most users respond within one hour of notification.

Opening rate of notification is 90% as compared to 40% for emails.

– source: Moengage

Some major drawbacks of m-commerce

There are not just good things that make up the difference between e-commerce and m-commerce. We also have some drawbacks of m-commerce. It has been criticized for creating a consumerism culture, addiction, impulse buying and non-sustainable practices.

Culture of consumerism

Another drawback of mcommerce is the increase in the culture of consumerism. Individuals have increased the spending on shopping. Users spend more than 4 hours on the apps. This has resulted in a situation where over 70% of the sales have started coming in from mobile apps.

Over 70% of the sales have started coming in from mobile apps

– source: eMarketer

Addiction of m-commerce

As we have discussed earlier, users are spending more than four hours each day. This leads to addiction of m-commerce. One big difference between e-commerce and m-commerce is the addictive nature of m-commerce. It can lead to eye fatigue, strain injury to hands and wrist and even psychological disorders.

Impulse buying behavior

Since users can take our their smartphone any time, any where, it is very convenient. However, this convenience comes at a price. It increases the chances of unplanned buying. Impulse buying refers to the act of sudden purchase of a product or service based on whim or impulse. Mcommerce apps are deliberately designed to increase this type of behavior through constant notifications, UI design, promotions and other selling tactics.

Non-sustainable consumption

One outcome of increased buying and impulse buying is that we are buying more than we need. This puts immense pressure on the already strained environment. We are generating waste at an alarming rate. Instagram and Snapchat culture has made it fashionable to have more and more apparels, accessories and paraphernalia. All of these have some carbon footprint and we are increasing our footprint as we increase our level of consumption.

The post Difference between e-commerce and m-commerce? appeared first on Management Weekly.

Investing in stocks is a tricky business. You do it wrong and you lose money fast. You do it right and you can laugh your way out of the bank. Companies like Apple and Microsoft have been some of the fastest-growing companies riding on the wave of the third industrial revolution. If you invested in these companies in the 1980s, you would have had a multi-bagger in your portfolio. This is the reason people are looking into the best ai companies to invest in.

Artificial intelligence could be the next big thing. Companies that rely heavily on AI can offer good returns for the shareholders as this technology matures. Although there is no guarantee, however, research shows that companies that rely on technological innovations typically offer higher returns.  

Methodology

We employ an aggregated information approach for preparing the list of top companies to invest in. We used the information from six different sources to prepare the initial list of companies. Then we filtered the list based on the AI centricity of the company. We eliminated companies that are not primarily based on AI. We used the company homepage and the latest annual report to ascertain the AI-centricity of the firm. Then we tallied the number of sources citing these stocks as value AI investment. 

The top 3 AI companies to invest in 

1. NVIDIA 

2. IBM 

3. Amazon

NVIDIA (NVDA)

Nvidia is an American company known for its graphics cards or Graphical Processing Units (GPUs). The homepage of their website clearly demonstrates their commitment to artificial intelligence. The title of the page reads ‘Artificial Intelligence Computing Leadership from NVIDIA.’ NVIDIA has invested heavily into technologies that can drive the artificial intelligence platforms of the future. The tech-savvy readers may be aware that AI can benefit immensely from the use of GPUs. GPUs are far more optimum for AI type of computation when compared to CPUs or Central Processing Units. A modern GPU may contain over 1500 to 5000 processing cores while a CPU typically has about 4 to 12 cores. Also, the manufacturing of such sophisticated hardware has very deep dependencies. You can’t have an overnight competition from new entrants. Remember Porter’s five forces? 

There is a strong reason for the inclusion of NVIDIA in most of the lists presented above. Firstly, hardware is an essential player in any digital transformation. If we look back to the 1980s and 1990s, Major hardware players saw immense growth. The steam story can be repeated in this industrial revolution as well. NVIDIA is one of the largest companies in the world that is focused on making hardware for AI computing. Therefore, it will enjoy a decent market share in the foreseeable future.  

IBM (IBM) 

IBM is one of The Pioneers who jumped on the AI bandwagon. They have been marketing their AI supercomputer, which is the IBM Watson for over two decades now. It got a lot of media coverage when it defeated human players in the game of Jeopardy. However, it has now evolved into a full-fledged enterprise AI solution.  

On one hand, we saw how NVIDIA plans to create value through AI hardware products. On the other hand, IBM positions itself as a service company. IBM offers a host of software as a service or SaaS packages for AI-based applications. Some of the key offerings are: 

1. IBM Watson Assistant – multichannel communication solutions 

2. IBM Watson Discovery – data analytics platform 

3. IBM Watson Natural Language Understanding – text analytics 

4. IBM Watson Speech to Text – transcription services 

5. IBM Watson text to Speech – TTS engine 

6. IBM Watson Knowledge Studio – domain specific knowledge 

7. IBM Watson Language Translator – translation engine 

8. IBM MaaS360 with Watson – cloud unification 

They have built some specific industry or domain-specific tools and solutions. It encompasses everything from IT security to marketing automation.  

Amazon (AMZN) 

Although IBM was one of the highly revered companies of the 20th century, Amazon commands a similar respect today. We can safely say that it is not a startup anymore. Amazon has matured as not the largest marketplace in the world, but also a leading cloud service provider. It offers a host of AI as a service over its AWS or Amazon Web Services platform. Secondly, Amazon is also one of the leaders in AI assistants. Amazon Alexa is the third most popular virtual assistant after Apple Siri and Google Assistant. 

When we specifically talk about the enterprise AI solution, the first Amazon product that comes to mind is Amazon SageMaker. They try to differentiate it from other solutions by making it easier to build and deploy. This platform promises easy deployment of ML algorithms. They offer solution-specific algorithms related to Business process centers, marketing, document processing, fraud detection engines, media intelligence, data analytics, DevOps, and business transformation. 

Are there some other AI companies to invest your money?

Of course, there is no dearth of good investment opportunities. Firstly, this list focuses only on American companies. There are several AI companies located in different places in the world. Also, many new AI startups are coming up every day. It requires higher due diligence to invest in these smaller companies. You may need to closely follow the latest IPOs to catch these investments early. We shall be updating this list with such opportunities soon. However, in this list, we have only included the top three companies that are relatively safer bets. These companies are also diversified into many different business domains. This diversification of business means a lesser risk for the investors. Please feel free to share your thoughts in the comment section below.

The post Top AI companies to invest in 2022 appeared first on Management Weekly.

You are surely missing out on a lot of potential customers if you have not listed your business on Google Maps. But we are here to help you. Just follow this simple 3 step guide on how to get business on google maps. After that, we have the 4th step where you start using all the benefits of this amazing platform. Just sit back and enjoy the flow of new customers. You may add more details, photos, website, and other information after you list your business.

Step 1: Visit Google My Business 

The first thing to keep in mind is whether you want to create a fresh Google account for your business or use the current one. Here is a very brief comparison of both options: 

Once you have decided whether you want to use the same account or create a new one, you may proceed further. You can either use this link to create your business on Google and Google Maps. It will open in a new tab so that you can switch back to our article for reference. Alternatively, you can search for Google My Business and click on the official link from Google.

Step 2: Claim your business

The next step is to search for any existing businesses. If you already have your business listed then you do not need to create another listing. You can simply select your business from the search results here and add your location to the map.

If you cannot find results for your business or store, click on ‘Add your business to Google. 

Next, you can add a location for the physical store. If you only offer online services and do not want people to visit your place, please select No. 

After searching for the name, you can add the location of your business or store

Select Yes to add the location of your business to Google Maps. Ensure that you put the correct and valid location as it will be verified later.  On the other hand, you may be offering services like home cleaning or online classes. In that case, you can click on no, so that people do not show up on your doorstep. This option is also valid for businesses like cloud kitchen who only have delivery facilities. However, they may not have any dining options.

Once you put your postal address, you shall be asked to drag and position the location pin to your address. Take your time and hit the exact spot. You do not want your visitors to be directed elsewhere. Once you are done, you have the option to choose whether you only offer location-based service. In that case, you can choose to cater only to people in your area. Otherwise, your business will be shown to people outside your neighborhood.  

Finally, it will ask you to add your phone number and website. Once you are done with that, you are all set for the verification process. At this stage, if you want you can also sign up for a free web page from Google. Otherwise, you can use your own website(recommended for most businesses). 

Step 3: Verify your business location 

Now you have to add your official name. You will receive a postcard with a verification number. Google will send you this postcard free of cost. From your end, you have to ensure that all the details you have provided are correct. 

Enter your name. Once you are done with this, you may select ‘Mail’. On the final page, you shall be asked whether you want to be verified by mail or other means. Mail is the best and simplest method of verification. 

Usually, it takes up to 14 days for this postcard to arrive. Be rest assured that no one else can use it. You have to enter it to verify your account. All the details and verification links will be provided on your card. If you have any doubts at any stage, you can contact Google at their toll-free number. This will be provided on the card itself. You may also ask me any doubts in the comment section below. Since I have done this entire thing a couple of times, I can surely help you with that. 

Step 4: Track the leads and also update your information from time to time 

You can start adding information for your business even before your verification process completes. For instance, you may add menu items for dishes that you serve at your restaurant. Or you can add some other details. For example, you may add photos of your reception and rooms if you are a hotelier.  

Also, do not forget to add enough descriptions for your products and services. People search for specific keywords when they do their search on Google or Google Maps. You do not want to miss out on these high-value keywords. Maybe you can enroll in some social media marketing courses to hone your skills as well. 

Why should you put up your business on Google Maps? 

1. Increase the footfall into your store or business. 

2. It will potentially lead to higher revenue and higher profits. 

3. Diversify your customer base. You may tap into a new segment of customers who come to your store via this route. 

4. Even if you have a website, you should still list on G Maps. Localization is very useful to feature on top of the search results. 

5. People who search for your product on maps have very high intent. High intent reflects that they are ready to make purchase immediately.  

The post How to get business on google maps? appeared first on Management Weekly.

As a researcher or analyst, one needs clarity regarding the pros and cons of a sampling method. Therefore, systematic sampling advantages and disadvantages will help you choose this sampling method for your study/analysis. Here we shall look at the various aspects of systematic sampling. Additionally, we shall try to understand it from different perspectives.

An introduction to Systematic sampling

Definition of Systematic sampling 

Systematic sampling is a common data gathering procedure. In this type of sampling, we collect the elements from a population-based upon a probabilistic collection method. In plain English, this means that we select the elements spaced out at some fixed interval.  For example, please see the illustration given below. Here we select every 2nd T-shirt from the lot. This is an example of systematic sampling. We could have even chosen every 3rd T-shirt from the lot and so on.

Characteristics of Systematic Sampling 

Systematic sampling is easier to understand and implement. It also makes the data collection more robust compared to convenient sampling. However, it differs slightly from simple random sampling. In simple random sampling, all the samples have got an equal probability of being selected. However, in systematic sampling, we do not have that. In systematic sampling, only each of the elements has an equal probability of getting selected. 

It can also be understood in terms of a three-step process. In the first step, we calculate the total population. In the second step, we asserted the sample size that is required for the study. Then we divide this population by the sample size. Therefore, we get the interval for collecting data. In the third step, we start collecting data from the nth individual in the population. This is the same end that we had calculated in the previous step. 

Where do we use systematic sampling? 

When the data is of higher quality and it is reliable, then we may go for systematic sampling. Therefore, we should avoid systematic sampling in situations where data could be manipulated. This will lead to erroneous results and also it will give misleading data visualization at the end of the analysis. 

We use systematic sampling in research to avoid any kind of data selection biases. For instance, data that has been collected using a probabilistic method, like systematic sampling, lead to a higher quality of raw data.  

What is an example of systematic sampling?

Let us consider the monthly output data of a small garment factory. We may be interested in the total units of a product made in different months. Now, let us say, we want to do a systematic sampling of the production. Consequently, we select monthly data for every third month (Sample 1 as shown in the table below). This would be an example of systematic sampling. On the other hand, we could also collect data for every six months (Sample 2 as shown below). Even that would be an example of systematic sampling. However, if you randomly select months 1,3,4, and eight (Sample 3). Then it is not a systematic sampling. 

Systematic sampling advantages and Disadvantages 

Advantages 

1. The first advantage of using a systematic sampling is that this type of data gathering procedure is fairly simple. 

2. At times, data collection is done manually by the researcher. In that case, it makes sense to have a systematic sampling as it eases the data collection process. 

3. It eliminates the biases of data collection present in a non-systematic method. 

4. It has a lower risk factor as well. 

5. This method is also useful to create a good distribution of the members of a population. 

Disadvantages 

1. One of the biggest downsides of systematic sampling is the poor quality of sample if the population has periodicity (Iachan, 1982)

2. Therefore, in a population that does not have randomly occurring data, it cannot be administered. 

3. It also introduces some risks of manipulation by researcher. 

4. Another disadvantage of this method is that we based our assumption on a determinate (measurable) population size. In a lot of situations, we may not know the size of population or it may not be determinate. 

5. In the pecking order of the sampling techniques, it is not the best one. Simple random sampling is better than systematic sampling. Also, stratified random sampling is better than simple random sampling. 

What is systematic judgmental sampling? 

Systematic judgmental sampling is yet another method. It combines the techniques of systematic sampling as well as that of judgement sampling. However, this type of sampling should be avoided. It introduces the errors from both these types of samplings. Therefore, we should avoid it.

The post Systematic Sampling Advantages And Disadvantages appeared first on Management Weekly.

Data gathering procedure is a critical step of any research or analysis. We have to ensure that we understand this important step so that our analysis is simpler and our results are also meaningful. Here in this article we discuss the different types of data gathering procedures and when to use which method.

What is data? 

• Facts 
• Figures 
• Statistics 
• A recording such as audio or video 

Data is a stored representation of information. Data is purpose-driven storage of information usually in a specific format conducive for retrieval and usage.  

In simpler words, data can be facts, figures, statistics, or certain other values. These could be numbers, text, videos, audio, or any other format. Usually, a single value of data represents a variable. For example, we can have data on the salaries of employees. On the other hand, if we take the salary of an individual from our employee database. Therefore, a datum is the singular form of data. Usually, when we talk about data, we’re talking about raw data. You can read more about examples of raw data in our other article. 

Why do we need to gather data? 

It’ll add ring is an essential step of the analysis process. It is one of the primary steps; therefore, it has a lot of implications in the later stages. These are the following reasons why we need to gather data: 

1. Allows us to store information. 
2. It forms the raw material for data analytics. 
3. It can also help us understand the context better. 
4. Gathering data will help us answer certain questions. 
5. Data gathering also helps us monitor the progress. 
6. Data gathering helps us in data visualization

Major types of data collection procedures 

We divide data gathering or data collection procedures into two broad categories. The first category is all primary data collection. On the other hand, the second category is called secondary data collection. 

When we are gathering data in the form of primary data, then the indulge ourselves in the collection process. We can use various different methods to gather data from appropriate sources. It is also important to understand the difference between sample and population.  

Secondly, we can also sample data from some source that includes a vast amount of data. For example, let’s consider the financial data. One way to collect the profit and loss information for a firm could be to write a mail to the company and seek this information from them. In this case, you are collecting the information directly from the source. Therefore, this is an example of primary data. However, if you want to collect data from 20 different companies, this mailing process can quickly become very cumbersome.  

It makes sense to collect the profit and loss data from a database, which includes the profit and loss statements of many different firms. Therefore, it makes sense to do secondary data collection in this case. You may use financial data sources like Bloomberg. Therefore, you would save both time and effort and collect the entire data required for your analysis within a short period of time. 

Primary data gathering techniques

One of the most common family data collection techniques is surveys. Surveys are a set of structured questionnaires that you provide to your respondents. These respondents fill up these surveys and then return their responses. Another type of primary data collection technique could be interviewed. When we collect data through an interview, we’re collecting qualitative data.  

Some other types of primary data collection techniques are user polls, focused group discussions, Delphi techniques, as well as ethnographic studies. A relatively new technique called netnographic technique is used as data-gathering produce where you collect data from websites and blogs by actively observing people’s behavior. 

Secondary data gathering techniques 

There are many ways to collect secondary data. Most notably in the world of finance research, we use financial statements and financial reports from companies. We may also use different financial databases for our information. Also, marketing managers may use different types of marketing surveys and reports that are available online.  

Finally, some other types of data that could be useful from secondary sources could be the data that is available on a company’s website and data that is available on the Internet. When you collect data from the Internet which is already present on the websites, then it is a type of secondary data collection. For example, you may want to collect data from the website by running some script that downloads text, images, etc. This is called ‘scrapping’ and it is a type of secondary data gathering procedure. 

Best practices for data gathering 

There are certain best practices for collecting data. We shall discuss some of them below. 

1. Even before you start gathering your data, you should ensure that you have the end object. If in your mind. This will help you gather only the relevant data. 

2. Secondly, it is also important to define your population correctly. At the same time, you should also ensure that you have defined your sample as well as the sampling technique before you collect the sample. 

3. When you do primary data collection, you’re out in the field to collect data. In these situations, you should ensure that we reduce the researcher’s bias in the data collection procedure. 

4. It’s also prudent to have a data-gathering system that gives you a sample that is a good representation of the population. 

5. Moreover, when you are collecting data from secondary sources, you still need to maintain rigorous standards. You should again maintain the same level of discipline and professionalism to ensure that the quality of data is good. 

6. Whenever you’re gathering data from secondary sources like databases, you should always note down the specifics. You may write down the queries or the type of filtering criteria that you have used to collect information. 

7. Once you have the raw data collected, it is also important to make backups of the data. If your data is already digital, it is fairly easy to copy it to another drive. Alternatively, you could also upload it to the cloud. However, if your data is in a non-digitized form then you may want to digitize it to make a backup. 

With the advent of Gen AI, there are some major changes that have happened in data collection. One needs to be aware of synthetic data, data transparency and privacy and data governance.

The post Data Gathering Procedure – 2026 guide appeared first on Management Weekly.

Data goes through different forms before it gets converted into useful insights. We take a look at examples of raw data. This will help us appreciate the underpinning raw material of analytics. It will also help us in understanding different types of data and how they have to be worked upon. This is the first step in using data analysis for decision-making.

The raw data 

Raw data is a collection of facts, figures, and other information that is available for analysis. This data is called raw because it has not been worked upon. Overall, there are three stages of data in analytics: raw data, processed data, and analyzed data.

Data collection 

The first step in data analytics is the collection of data. The data that we collect is typically raw data. We can have different sources of raw data. For example, we can collect data from primary sources like interviews or surveys. However, we can also have secondary sources of data like government, statistics, and some publicly available data. 

In certain cases, we may be interested in analyzing the data that is present in a database. In such cases, we have to do the collection of data by running SQL queries. This data is also an example of raw data. 

Why is raw data important? 

Raw data is important because it is the first step in the journey of data from potentially useful to a useful form. A lot of attention has to be given before the data is collected. Once the data has been collected, we are limited by the quality of data that has been collected. If the data collection is not done properly, then we are stuck with less-than-optimal raw data. This could lead to less-than-optimal results at the end of the analysis process. 

Secondly, raw data is also important because it can offer some insights on its own. There are times when analysts can find patterns in the raw data itself. One of the ways in which pattern can be found in raw data is by running specific queries in SQL that gives a summary result of the database. However, more commonly raw data can be analyzed in Excel itself. Sometimes we are able to make sense of the data by looking at the tables and their values in Excel. 

Thirdly, raw data is also important as a trend-setter for later stages of analysis. Care should be taken at this stage to represent and record the data effectively. Correct data entry will help the process later on by eliminating the possibility of errors. Some of the errors of data visualization occur during the raw data management stage only.

What are the types of raw data? 

There are many different types of raw data. However, for our purposes, we shall define raw data in terms of the data collection procedure. Broadly, we can say that there are two types: qualitative data and quantitative data. 

Examples of Raw Data 

Survey of weights of students in a classroom 

In this example, we look at the distribution of the weight of students in a classroom. In this example, the weights represent the raw data. This data can be captured or arranged to produce a meaningful output. For example, we can find the average of all the weights to find the average weight of a student in the classroom. We can also find the maximum weight of a student and the minimum weight of a student from this table.

An example of Sales Data

When we collect data from a small sample based upon a small number of variables, that data is typically small data. Although there is a lot of impetus in recent years on big data, even small data could be quite useful for analysis. It is much faster to collect this type of data and it is much easier to process this data. A smaller data set can also be processed manually by the analyst. The table below is one of the simplest examples of raw data. 

Raw audio data from interview recordings 

 
 
One of the examples of raw data is the audio files that we get from recording interviews. Please check an example of a recorded audio interview below. This reference file is taken from the US Library of Congress.
 

Source US Library of congress 

Raw video data 

Another example of raw qualitative data is raw video data. Video data can also be available in different formats and sizes. Usually, there’s no data that is captured from a camera is high bitrate. It needs a lot of space for archival. However, raw data captured from smaller or consumer-grade video recorders and smartphones may already be compressed.  

Social Media data as raw data

Social media analysis is another important way to understand market patterns and get some user insights. However, it is also more challenging due to the following aspects:

1. Social Media Data is an example of raw data that is unstructured. Unstructured data is difficult to analyze.
2. It is also difficult to capture social media data except for platforms like Twitter and Youtube.
3. A lot of social media data may not be useful. Therefore it takes more effort to find useful patterns.

What are the common pitfalls with Raw Data? 

1. More data is not always better 

2. Starting with raw data rather than with hypothesis 

3. Trying to find answers in raw data 

4. Not handling raw data properly 

The post Examples of raw data – for analysis and research appeared first on Management Weekly.

The graphical representation of data enables us to present information in such a way that is not possible by any other means. Research has shown that data visualization increases the acceptance of information compared to other approaches. Therefore, it is essential to understand what is the best way to visualize data.

What is the best way to visualize data? 

Although there is no right answer to this question, the best way is dependent on a number of factors. However, you can use a step by approach to finding out the best way to represent your data. Answering these questions will give you some clarity. 

1. What is the type of information? 

This is the primary question you should be asking before jumping into data visualization. In order to do this, we need to look at the information that we have at hand. Firstly, information can be broadly categorized into two types: real-world and empirical data. In the first case, we attempt to model real-world objects. On the other hand, the empirical data captures a specific aspect of the real world or their interaction with one another. 

For instance, real-world information can represent a physical object such as a ball kept on a table. We can see the dimensions and position of real-world objects and represent them visually. Although, we may choose to reduce or increase the scale (size of the object) as per the need. On the other hand, we may also have conceptual objects like forces acting on the ball. Although we cannot see these forces, we can objectively measure them and represent them visually. 

The second type of information is empirical data. When we have to understand the pattern in individual cases, we have to take a large number of samples. We may capture this type of data numerically (quantitatively), non-numerically (qualitatively), or a mix of both (mixed data). 

2. What are the objectives? 

There are four broad objectives of data visualization (Berinato, 2016): 

1. Illustration of an idea 

2. Idea generation 

3. Discovery through visual representation 

4. Representation of simple relationships between variables

Illustration of an idea 

We need to represent some ideas commonly in the presentations. In the business context, it can be the representations like BCG matrix or ‘customer journey mapping.’ Although it may take time to develop such visualizations, it helps in presenting your information. It would also save time for your audience since they can absorb the insights more readily. 

Idea generation 

Idea generation is the real-time visualization technique. For instance, you may use a method like mind-mapping to take a note of the idea generation in a session. Firstly, such representation helps us in keeping track of a lot of information that is difficult to remember. Secondly, it also helps in organizing the complexity of the information. This shall help in making sense of data. Thirdly, it may help us in quickly going through the discussion later.  

Discovery through visual representation 

Sometimes, visual representation gives us totally new insights. An example of such representation is a bar chart that compares the average ticket size (purchase or basket size) of female vs male customers at a store. We can use both a table and a bar chart in this case. However, most viewers will process a bar chart more easily. Taking it a step further, you may develop an infographic representing female and male figures to make it more appealing.  

Representation of simple relationships between variables

Finally, we come to the 4th type of data representation. In this type, we try to represent the relationship between variables. A univariate data analysis looks at a visual representation of a single variable. For example, if we want to visualize what amount of salary is more common than others, then it is analyzed and represented as a univariate. On the other hand, if we want to see, how salary varies with the level of education, then we have to represent that as a multivariate.  Managers often need to use multiple regression to find patterns in data.

Therefore, your objective will define the type of visualization method that you will use. These methods. Can range from fairly simple ones like bar charts to a more advanced and technically challenging one like infographics. Our next question will help us answer this choice. The choice of a particular visualization method is dependent on the target audience. 

3. Who is your target audience? 

When you are trying to visualize the data for a less educated audience, then your approach should be to represent the information in the easiest way possible. Your driving concern should be the representation of data that is both easy to read and easy to understand. One of the ways of doing this is using a summarized version of data. 

On the other hand, if your target audience is well informed then you may want to provide them with advanced details about the data. However, the goal of the analyst should still be to summarize the information in a digestible form. You should try to avoid using too many graphs and too many charts that can digress you from the message you want to convey. 

4. Data Visualization and Presentation 

Once you have made the choice of the type of data visualization, you have to choose the method. There are various methods to represent the data visually. However, the best way to visualize data will be dependent on the factors that we have discussed above. The next section brief you discuss is all the different options that you have.  

We should keep in mind certain factors while we visualize the data. 

1. The number of variables in the data. 
2. The number of cases. 
3. The type of data. 
4. The data transformations
5. Choosing an appropriate data visualization technique
6. Deployment of the visualization

Some points to keep in mind

The most common representation of data is either univariate or bivariate. However, in certain instances, we want to see the relationship between three different variables.  

Each data analysis requires a certain number of minimum cases to be statically significant. The same logic applies to the visualize data as well. You should ensure before data visualization that you have statistical significance in the data. For example, imagine that you are trying to show the group differences between two market segments. You should first check for the group differences using some kind of methods like T-test or ANOVA.  

If you have got two types of data. The first type, being categorical data and the second type being numeric data. In this case. It is more common to keep the categories on the X-axis and keep the numeric data on the Y-axis. For example, if you’re trying to plot a bar chart of the sales with respect to the days of the week. In this case, it is much better if you keep the day of the week on the X-axis. Consequently, you have to keep the sales that happen each day on the Y-axis. 

Sometimes you may have to do data transformation to make the data visually easier to understand. For example, consider the case that you want to show the number of active cases of covid19 with respect to time. In this case, if you use raw data, then the graph that you will get will not be a good representation of the information. It will be more difficult to extract valuable insights from this chart. The reason for this is that the case is group and reduced in an exponential fashion. Therefore, our raw chart may not be interpreted correctly. On the other hand, you may apply a log transformation to the number of cases. This will reduce the amount of kurtosis in the graph and make it easier for people to understand how the cases grew over time. 

Top 5 innovative visualization methods in 2021

Understanding the different types of visualization helps us in presenting the information in the most conducive way.

1. Infographics

Infographics are an incredible way to visualize data. It makes the representation both appealing as well as alluring. Earlier, you needed to be a graphics designer to create infographics. However, there are many easy-to-use tools that have made creating infographics a breeze. Therefore, it is on the top of our recommendation list for the best way to visualize data.

2. Heatmap

A heatmap is a visualization that uses color overlays to represent variation. Traditionally heatmaps were used to show which areas of a city had more crimes. However, it can be used in any type of data. You may even use Microsoft Excel to create heatmaps easily.

3. Geographical visualization

Geographic visualization is another great technique. Similar to the infographics, it was challenging to create geographic/cartographic visualization. However, with the advent of software like Tableau, it has become much much easier. You can overlay your insights on the map of your country, city, or the world in a matter of minutes.

4. Treemap

If you are trained in statistics, you can readily think of treemap as an extension of decision trees. It is a 2D representation of hierarchical data. The data is nested in the form of shapes like rectangles. Each rectangle can be clicked and you can see the data distribution inside that rectangle. Again, this type of visualization can be best made using dedicated data visualization software. However, with a little bit of effort, you can create the same in Powerpoint or Excel.

5. Streamgraph

A streamgraph is like a stacked area chart. However, it is created around a horizontal axis. This makes streamgraph look like a flowing stream. Usually, different colors represent different variables. It is the best way to visualize data that is temporal. Temporal or time-varying data changes over time. This trend could be represented well in a streamgraph.

A list of top 20 most common visualization methods

The post Best Way to Visualize Data – a guide for 2021 appeared first on Management Weekly.

Volkswagen group was a highly regarded European automaker. It was seen as a major rival of the Japanese giant Honda. However, things fell apart very quickly in 2015. The stocks tumbled over 20% in a single day when their infamous Dieselgate Scandal came out. This is a stark reminder of the ethical issues of technology.  The often-overlooked aspects of ethical decision-making may lead to expensive blunders.

What are the ethical challenges today? 

Ethics is the set of guiding principles that provide a way of acting morally during decision-making. Some decisions may not have any possible implications. However, we face ethical dilemmas every day at the workplace. Your choice of particular strategy may fetch your firm higher revenue but may mean lower earnings for small retailers. Similarly, what would you do, when you learn that your database has been hacked. How would you disclose this information to your millions of users? Or should you hide it till it gets noticed? 

⚡ Ethics is the set of guiding principles that provide a way of acting morally during decision-making.  

In this section, we uncover some of these ethical challenges. These ethical issues of technology are faced by firms of all sizes. The five most common challenges are discussed below.  

1. Data Privacy 

2. Data Misuse 

3. Misinformation 

4. Accountability, Responsibility and Liability of actions 

5. Intellectual Property 

Data Privacy   

Data privacy is one of the major challenges faced by contemporary firms. European Union defines data privacy as well as keeping data protected from unauthorized access. They have provided a comprehensive guideline for the firms operating in their geography. 

Data Misuse   

Data misuse is another major challenge faced by many firms. There is a growing mistrust among the consumers that their data may be used in contexts they have not approved. Also, many believe that companies had historically not disclosed all the data they were collecting through the devices. 

Misinformation   

Thirdly, there is another major challenge – misinformation. The first two problems are associated with the breach of access or breach of usage terms of the data. However, misinformation is about the creation of data with wrong intentions. These intentions can be relatively benign like promoting a new product. However, they could also be harmful ones like fake news to incite racial crimes. 

Accountability, Responsibility, and Liability of actions   

The fourth major challenge is that of ownership of the problems. Accountability is owning the mistake that has occurred. As we have discussed the Volkswagen case in the opening of this article, the firm was aware of the cheat device installed in their diesel cars. Responsibility is about knowing what is expected from you and working towards completing that goal. Once Volkswagen realized their blunder the knee-jerk reaction was to replace the top management. However, they took up the responsibility as well as have been working to improve the CSR at Volkswagen. Liability is the legal obligation to pay for any problems that occur. This could be due to negligence or maleficence.  

Intellectual Property   

Finally, we have the challenges related to intellectual property. On one side of the equation, we have the firms. It is expensive and effortful to manage intellectual property. Firms like IBM and Samsung file thousands of patents every year. In 2019 alone, about 333k patents were filed in the United States. On the other side of the equation, we have the consumers or users of technology. Piracy is a less-discussed but very important aspect of the ethical issues of technology. Companies lose a good chunk of their revenues because of piracy. This is particularly true of companies that deal with digital products like software, music, and video. 

Ethical Analytics Model 

We can divide the issues pertinent to technology into three categories as per Laudon and Laudon: 

1. Ethical issues 

2. Social issues 

3. Political issues 

These issues are interconnected. This provides us with a base to assess the decisions.  

The ethical analysis requires us to question certain basic assumptions. Here are ten guiding principles to make decisions. These shall come in handy when working with the ethical issues of technology. 

The principle of non-maleficence 

This principle states that do no harm to others or primum nil nocere. This may seem to be fairly simple. However, it is difficult to practice it in the face of decision dilemmas. It becomes even more challenging to follow this when it comes to situations where doing no harm to others may entail some harm to oneself.  

Beneficence 

The primary goal of any business is to maximize the stakeholders’ happiness. The actions taken by the management should Increase should benefit the customers, employees as well as shareholders. While the principle of nonmaleficence looks at averting wrongdoing, beneficence looks at actively helping others.  

Ensuring due compensation 

Firstly, when we talk about technology, we need to credit the creators. This credit may include monetary as well as non-monetary compensation. At the same time, we should also ensure that any intellectual property we use should be credited. Some information or code may be available under copyright-free licenses like Creative Commons. We must adhere to the requested format for such reuse. 

Undue advantage 

There are certain situations for him. The organization may benefit from taking undue advantages. For instance, a company may cheat the pollution control technology to produce steel at a cheaper rate than the competition. However, if everyone starts doing that, we will have massive problems with pollution. 

Value maximization 

Sometimes choices could be difficult because different choices may seem appropriate. In this situation, it is important to consider the choice that maximizes the value for the most number of people. In this case, we can rank the different choices that we have in terms of the overall utility it provides to the different stakeholders. 

Risk minimization 

When we are choosing between different alternatives, another thing to consider is to choose an option that reduces the risk. For instance, let us consider that a company is trying to make a new vaccine for a virus. In this case, if the firm is able to market the virus fast, then it maximizes its own profit. However, if proper testing is not done, then it may increase the risk for the people who take the vaccine. Therefore, it is important to consider this aspect and not hurry in the release of the vaccine. 

Respect for privacy 

Data is the new oil. Organizations use this data to run their businesses smoothly. However, some of this data may be linked with the personal information of the individuals. It is important to value other’s privacy when it comes to information usage. Firms should also ensure to have checks and safeguards against data misuse and theft. You can gain some interesting insights from quotes on business ethics collected from different thought leaders.

Concerns Around Ethical Issues of Technology

In this section, we have tried to quantify the ethical issues of technology. Some of these issues may have been historically important. However, the changing face of technology with the advent of AI and automation has brought about some new issues in this decade. Some of the critical ethical issues today are listed below. 

1. Data usage without consent 

2. Privacy of sensitive information 

3. Misinformation. 

4. Operate infringement. 

5. Deepfakes 

6. social media trolling 

7. Lack of accountability. 

8. Genetic engineering and genetically modified crops. 

9. Unregulated players in the fintech industry. 

10. Human biases in AI. 

11. Data manipulation. 

12. Misuse selling of private data. 

13. Weaponizing new technologies. 

The post Ethical Issues of Technology: a definitive guide appeared first on Management Weekly.

The acronym KDD stands for Knowledge Discovery in Databases. KDD process in data mining is an analytical method for finding useful patterns in data for practical applications.  

What is KDD? 

Knowledge discovery is one of the most powerful applications of information systems in business. We use KDD to convert useless and often overwhelmingly large data into useful information that can help in decision-making.  

Essentially we use data to answer some of the business problems. Consequently, we these insights to meet organizational objectives.  You may read in more detail about the significance of data analysis in decision making in our other article. Sometimes data mining is synonymous with KDD. People may use these terms interchangeably.

What are the expectations from Knowledge Discovery? 

KDD or data mining is the “process of extracting valid, previously unknown, comprehensible and actionable information from large databases and using it to make crucial business decisions”. 

The knowledge discovery process in data mining must fulfill these expectations: 

Non-trivial  

For instance, if you discover that people who buy luxury apparel are richer than people who don’t buy them. However, if you are able to find out specific life events that may trigger the purchase of luxury products then that knowledge discovery is non-trivial. 

Previously Unknown 

Secondly, it is also important to discover something new. The analysis is worthless if it provides a piece of stale news to the decision-makers. For instance, your executives may already be aware that your products perform poorly in certain months. Merely finding the same result won’t add any value.  

Implicit 

This is another important feature of the KDD process in data mining. Essentially there are two types of knowledge – explicit and implicit. Explicit knowledge is often evident and clearly visible without much effort. However, implicit knowledge is embedded inside the data. Implicit knowledge includes the unwritten rules of processes, individual learning. Therefore, they are difficult to mine.  

Useful 

Finally, and most importantly, knowledge would be useless if it lacks practicality. It is also useless if it is not related to the problem at hand. Therefore, the analyst must sift through the output of data analysis to select only the information that is useful towards the business objective.  

Simplified Knowledge Discovery Process 

The contemporary approach of extracting knowledge is somewhat different from what is depicted in the traditional KDD diagrams. There is a greater focus on reusability. Analysts must not start the process again from scratch. Therefore firms put a lot of effort into developing data marts that take a lot of time initially. However, we can do queries and data extraction more efficiently. Also, we should do data cleaning and transformations in the initial phase itself. That eliminates the need for this process later.

1. Data Extraction, cleaning, transformation and loading process. In the first stage, we extract the data. Subsequently, we clean and prepare it for loading into the data marts. The dimensions are transformed into the required format before it is loaded into datamart.
2. Second stage is extraction of useful and relevant data from the datamart. Relevant data is extracted by running queries on the datamart. These queries provide us with a subset of the data in the datamart.
3. The third stage is to analyze the data. In this stage, the data is sliced, diced, summarized. We may also use the data to build statistical models.
4. In the final stage the output is represented in human consumable format. Typically, the business executives require insights in the form of tables, charts and infographics. This information can either be presented on-demand or in the form of customizable KPI (Key Performance Indicators) dashboard.

KDD Process in Data Mining

Databases 

A database is defined as “an organized collection of structured information, or data, typically stored electronically in a computer system” (Oracle.com). The most important concept here is the structured nature of the information. It is very difficult to make sense of unstructured data, especially when the dataset is large. 

However, databases have certain problems which inhibit their usage for data analysis. Firstly, it takes a lot of time to run queries* on databases. Secondly, Datamart is specific to business usage. Therefore, they are more efficient and they cost much lower to be implemented. Thirdly, required data can be scattered in different databases. If we have all the relevant data in one place, it makes the analysis both efficient and faster.  

🛈 Sidenote: Queries are like commands that enable us to work on the database 

Datamart 

ETL process helps in building data marts from data warehouses. A data warehouse is a collection of inter-related databases. There may be different sources of data. Operational data, transactional data, customer data are some of the examples of different databases that can be accumulated at the Data warehouse. The data goes through the ‘extraction’ stage where the data from different sources are collected. The ‘transformation’ stage ensures that data is cleaned, processed, and standardized. Finally, the data is loaded into the data warehouses.  

In some cases, these data warehouses may be an intermediate stage, and then it is again used to prepare the datamarts. On the other hand, sometimes data can be directly used to prepare the datamarts. These are also known as enterprise data warehouses or EDW. Therefore, the data warehouse is like a single source of truth, where did require data from different sources are collated and put together. However, we should not run queries on EDW directly. These tables usually contain a lot of data columns that may not be useful for analysis. Therefore, we create smaller, manageable, and analysis-oriented databases. These are called data marts.  

Relevant Data 

Once we have the required data marts in place, we can extract the relevant data from the datamart. This data can be used for analysis. Let us consider a simple example to illustrate this step in the KDD process in data mining. 

You are interested in knowing the top-selling products in a store. Therefore, you run a query to extract a table that contains the point-of-sale data. Each row containing only the product name, SKU number, product category, quantity, unit cost, and total cost. Thereafter you may analyze the data to get useful output from them.  

Pattern  

The pattern is nothing but a useful output you get from the analysis. We may use an appropriate statistical method to work on the extracted data. These patterns represent the trends in data. For instance, let’s say that you have got some data points for sales in our particular shop. In that case, the pattern of sales will reveal that the weekends perform better than weekdays. However, this kind of simple pattern is not very useful. The reason for this is that anybody, even without the analysis, can say that weekends are better than weekdays because there is higher footfall. Therefore, there will be higher sales. 

Consequently, we have to put in some more effort to find out patterns that are nontrivial. Therefore, we may have a deeper look at data and try to see that which of the weekdays are performing better than others. A further drill down in this data gives us that Fridays perform the best while Tuesdays perform the worst. This kind of insight would be much more useful because it gives us a starting point to design our marketing campaigns to boost sales. 

Knowledge 

The patterns are then interpreted as knowledge. There should be novelty and usefulness in this knowledge. This knowledge is essentially visualized from the patterns that we get. We can represent the output of data analysis both visually as well as textually. Visual representation is in terms of tables, graphs, charts, and diagrams.  

Subsequently, we make sense out of this data and try to answer the business problem that we had posed in the first step. Again, we need to ensure that the insights that we have developed through the knowledge discovery process are non-trivial, novel, and useful. Also, this is not the end of the KDD process in data mining. Further, we have to use this information to fine-tune our models and iteratively find certain solutions which are not available on the first try. 

Some advanced concepts in KDD 

Data Scrapping 

Data scraping refers to the process of collecting information from sources like websites and blogs. Companies may collect data from these kinds of sources for different purposes. This data is used for analysis. Typically, data scrapping is done using automated scripts or software.  

For example, let us consider that you want to understand why some of your products are selling better than others. Also, you sell the product both on your own website and on Amazon. You may want to scrap the reviews from Amazon’s website. You can collate all these reviews together and do sentiment analysis on your products. This can help you understand what are the reasons that drive the sales of well-performing products and what are the reasons that inhibit the sales of poor-performing ones. 

Data Stream Mining 

Data streams are sources of data where data is characterized by continuity. Unlike traditional data which is non changing, these types of data streams offer continuously changing data points. One of the common examples of this kind of data would be weather data. The second example of this kind of data could be data emanating from IoT sources.  

One of the biggest challenges with our data stream is the extreme volume of data. For instance, if you’re collecting data from an IoT sensor, the first step would be to minimize the burden of data. You may take an average of sixty readings in a minute and only store temperature data for every minute. This will not only help you reduce the storage requirement; it will also make analysis more meaningful as it may even out random fluctuations in the sensor.  

Social Media Data 

Knowledge discovery can also happen from social media data. Social media is set to be one of the richest sources of customer’s voices. However, this kind of data is difficult to procure. Moreover, it is also difficult to analyze. One of the distinctive characteristics of social media data is that it is unstructured data. People can share posts that are textual or it could be images or it could be videos or any other format.  

The non-standardized nature of this data makes it difficult to be analyzed using traditional data analysis software. You may use special software like Hadoop, Apache Storm, or Google BigQuery for this kind of analysis. However, not all social media data is unstructured and some analysis may be done directly on the content. 

The post KDD process in data mining appeared first on Management Weekly.