Big Data Analytics Assignment Sample by NAH UK

Introduction to Privacy And Security Considerations For A Customer-Facing Apartment Rental Application Assignment Sample

Task 1: Rent “price” investigation

Assumptions and Pre-processing

This investigation is based on the assumption that the provided dataset, apartment_for_rent_train.csv, is both representative of the broader rental market and of sufficient quality to support exploratory data analysis and model development. Initial data inspection revealed missing values, which were removed to ensure the reliability of subsequent analyses [6]. Several categorical variables, including Duration, Address, and Furniture, were converted into numerical format using label encoding to make them suitable for model input. Additionally, rental durations were normalized into numeric days to facilitate correlation analysis, allowing for consistent comparison across listings with different time-based rental units (e.g., daily, weekly, monthly).

This assignment sample explores rental price analysis for a customer-facing apartment rental application, focusing on data cleaning, missing-value removal, categorical encoding, and rental duration normalisation. An Assignment writing helper can support students in understanding exploratory data analysis, data preprocessing, and reliable model development.

Key Discrete Variables Affecting Price

To identify which discrete features most significantly impact rental prices, boxplots were generated for several binary apartment attributes, including Balcony, Elevator, Parking, Renovation, and Furniture. Among these, three variables—Balcony, Elevator, and Parking emerged as the most influential. Apartments featuring balconies demonstrated notably higher median rental prices, likely due to the added value of outdoor space, a benefit well-documented in real estate literature. Similarly, units in buildings with elevators commanded higher prices, particularly in multi-story structures where elevator access is associated with convenience and accessibility [4]. Lastly, dedicated parking was associated with increased rental prices, especially in urban environments where such amenities are often scarce and highly valued. Boxplots (see Figures 1–3 in the appendices) visually confirmed these trends, illustrating clear price distinctions based on the presence or absence of these features.

Correlation Between Rooms, Price, and Duration

Pearson correlation coefficients were calculated to evaluate the linear relationships between Number_of_rooms, Price, and Duration_numeric (rental period converted to days). The correlation between the number of rooms and price was moderate and positive (r = 0.487, p < 0.001), indicating that larger units generally command higher rental prices. The relationship between number of rooms and rental duration was weak but positive (r = 0.061, p = 0.023), suggesting that apartments with more rooms might be rented for slightly longer durations. Conversely, the correlation between price and rental duration was weakly negative (r = -0.056, p = 0.040), implying that short-term rentals tend to be more expensive on a per-period basis, possibly due to higher turnover and management costs. A heatmap (see Figure 4) visually supported these statistical relationships.

Predictors of High Rental Price: Role of Address and Other Factors

A linear regression model was built using 13 input features, with categorical variables such as Address converted via label encoding. The model yielded a strong performance, with an R² and a root mean square error (RMSE) of 12494, indicating a reasonably accurate fit to the data.

Based on model coefficients, the five features of top influence when predicting rental price were the Address, Number_of_rooms, Floor_area, Balcony, and Elevator. Address was the strongest predictor and reflected location specific effects such as neighborhood affluence, proximity to free public services or just desirability, in line with the conventional real estate assessment models. The price was strongly linearly related with floor_area, where bigger living spaces tend to be rented at higher rates. Number_of_rooms, Balcony and Elevator reaffirmed their roles in rental pricing which have been found earlier through visual and statistical analysis. These contributions were represented in a feature importance chart with help to understand interpretability of linear regression model.

Critical Evaluation

The main strength of this investigation comes from the fact that the model is interpretable. With the help of linear regression, it was easy to clearly see how each feature was correlated with the rental price [1]. Pre-processing of the data was also efficient including cleaning, encoding, as well as transformation steps that facilitated and ensured the reliability and robustness of the analysis process that was carried out [7]. But there are certain limitations to this. Giving the Address variable as a straight number may have oversimplified the complicated effects of location and perhaps a better way to encode the Address variable would be a variety of more sophisticated such as geographic clustering or natural language processing based embedding that provide richer contextual information [9]. Furthermore, the presence of multicollinearity amongst variables (i.e. Number_of_rooms and Floor_area), may have weakened the coefficient stability, and hence the model robustness as well.

Reflection

The approach taken provided meaningful foundational insights into rental price determinants overall. A clear understanding of the key variables affecting rent levels was easy because this combined a statistical analysis and an interpretable model. For future improvements, instead of simple linear models, more complex models like Random Forest or XGBoost could essentially be used to identify non linear relationship and interaction among features. This would also help generalize and have reliable model. Furthermore, the dataset could be enriched with further geographic or economic indicators from outside the data, for instance proximity to public transport or local income levels, which could substantially boost the theoretical explanatory power and the practical utility of the model.

Task 2: Storing data and possible solutions

Part 1: Relational Database Design

1.1 Overview of Design Approach

A relational database model was designed as per the requirements of data integrity, scalability, and query efficiency to address the present limitations of the flat file system to store the apartment_for_rent dataset. This makes storage more structured and remove redundancy and dependency anomalies by normalization. All of the dataset was decomposed into 4 tables: Tenant, Building, Apartment and Pricing. These tables are each a logical entity that is derived from the original CSV structure.

1.2 Table Design and Relationships

The demographic information of the Tenant is recorded in terms of Reg_id, Gender, Age, Children_are_welcome and Pets_allowed which is recorded in the Tenant table. It recognises physical characteristics on top of these in Building table which are Construction_type, New_construction, Elevator, and Floors_in_the_building. Now, this abstraction prevents duplicating identical building specifications in several apartments. The primary key is assigned directly on Building_id to which each unique building configuration belongs. Then, the Apartment table consists both of Ref to Reg_id (foreign key to Tenant) and Ref to Building_id (foreign key to Building) as well as some attributes like Floor_area, Number_of_rooms, Number_of_bathrooms, and etc of the interior. Finally, the Pricing table is separated with financial and temporal aspects (as Price, Currency, Utility_payments and Duration) with Apartment_id as its primary key to the rest of the tables.

1.3 Normalization and Justification

This schema is normalized to Third Normal Form (3NF). All attributes in each table depend solely on the primary key, and transitive dependencies are eliminated by isolating them into related tables [3]. For example, by separating pricing from apartment specifications, we avoid redundant data entries when prices change.

1.4 Sample SQL and Queries

To demonstrate practical implementation, SQL statements were written to insert a new apartment listing spanning all normalized tables and to perform essential queries. One query retrieves apartments priced at $1000 or less, that include an elevator, and are listed in USD. Another calculates the average price per currency. These queries ensure the model's applicability for real-world analytics and reporting.

Part 2: Scalable Solution

2.1 Assumptions and Growth Expectations

Given the apartment manager's plans to scale operations internationally, the current relational setup, while effective for moderate datasets, must evolve to support increased data volumes, real-time decision-making, and geographic distribution [2]. It is assumed that the dataset could grow to include millions of records, requiring faster data retrieval and processing across distributed locations [5] . In this context, proposing a scalable architecture utilizing big data technologies.

2.2 Proposed Architecture and Technologies

For distributed storage, cloud-based object storage systems like Amazon S3 or Hadoop Distributed File System (HDFS) provide fault-tolerant, scalable environments ideal for large datasets. To handle processing at scale, Apache Spark is recommended due to its in-memory computing capabilities, which provide significant performance advantages for iterative tasks such as model training and real-time analytics [10] . Spark’s DataFrame API supports SQL-like queries while maintaining scalability, making it a suitable tool for the apartment price investigation use case.

2.3 Real-Time Analytics and Alerting

For real-time alerting and analysis, a combination of Apache Kafka (for event streaming) and Spark Streaming (for processing) enables immediate detection of critical business events. For example, if the number of available apartments in a specific location with desired amenities exceeds a threshold, the system can automatically trigger notifications to the management or customers.

2.4 Comparison with Alternative Solutions

Compared to traditional RDBMS solutions, distributed frameworks like Spark and NoSQL systems (e.g., MongoDB or Cassandra) offer higher throughput and better fault tolerance under heavy workloads. While RDBMS excels in ACID compliance and structured query capabilities, it lacks the elasticity and scalability needed for globally distributed datasets. NoSQL databases, with their flexible schema and horizontal scaling, are better suited for the unstructured or semi-structured data that may be collected from customer-facing applications.

2.5 Conclusion

In conclusion, a hybrid architecture combining relational databases for core structured data and big data technologies for scalable analytics offers the most balanced approach. This strategy supports the current need for structured analysis while preparing the infrastructure for future growth and responsiveness. The proposed technologies and design principles align with best practices in big data system design, ensuring that the system remains robust, scalable, and responsive as the business expands.

Task 3: Considering Customer-Facing Application

3.1 Privacy Concerns in the Customer-Facing Application

As the apartment manager considers launching a customer-facing application to facilitate browsing and recommendations, several critical privacy issues arise, particularly regarding the handling of sensitive personal data. Three of the most pressing concerns are:

Data Collection and Consent: In the case that the application is likely going to collect names, contact details, preferences or any other personal sensitive data including financial data (e.g. payment details), this is something which should not be shared without permission. This could lead to privacy violations without proper consent or with transparency. Customers must be given the clear information about what kind of data is to be collected, the reason behind the collection process, how the same is used and how long it will be stored [8]. Collecting personal information must be performed in line with regulations like the General Data Protection Regulation (GDPR) and explicit consent must be obtained for this purpose.

Data Storage and Retention: Storing personal data over long periods certainly doesn’t reduce risk. Breach occurs when data is stored for long or without securing it. Moreover, with lacking retention policies, data can be put to unintended use, like to promote targeted advertisement or reveal to third parties without the right consent. Retention protocols have to be put into place to dictate how long various data will be held and when it will be erased completely securely.

Data Sharing and Third-Party Access: The risk comes from sharing of personal data with third parties like service integration, analytics, marketing, and many more in which unauthorized access or misuse of data is possible. Third parties who are the recipients of the data should be made aware to the users and explicit consent is popped in before knowing sharing the data. In addition, these third parties have to conform to data protection standards to avoid loss of data from their hands.

3.2 Mitigation Strategies

The following is a number of strategies that can be used to mitigate these privacy concerns:

User Consent Management: For sharing data and conducting research, one of the most effective mitigation strategies is to get that informed consent before collecting all that personal data. At the beginning of the usage of the app, users should clearly see a consent to explain why data is being collected, how it will be used and for how long [12]. In addition to this, users should easily be able to withdraw it at any time using the app.

Data Minimization and Retention Policies: Data minimization principle makes sure that the app collects only required data necessary for the app to work. In addition, a data retention policy must be created on the amount of time personal data is retained, and when it will be securely deleted. There should be regular audits in place for the legal requirement (e.g. GDPR) to be met.

End-to-End Encryption and Secure Data Storage: Finally, an end to end encryption should be used to ensure data gets encrypted throughout the transmission and storage procedure. This keeps the customer data unreadable for unauthorized users. Storage of data should be stored in secure solutions such as databases whose storage is in the cloud or servers that use strong security certifications such as ISO 27001 [11]. Security assessments should also be conducted on a regular basis, including penetration testing to determine vulnerability and address them.

Third-Party Vendor Risk Management: Data processing agreements (DPAs) should be set up with third party service providers (e.g. payment processors); the process would then mirror agreements with partners that are used to receive data. These agreements guarantee that third party’s are up to privacy and security standard. In addition, vendors should also be audited from time to time to ensure the compliance and when any security breach occurs, it should be reported quickly to users.

3.3 Conclusion

A significant concern with the proposed customer facing application is privacy. In order to effectively eliminate these risks, the apartment manager will have to put into place measures like informed consent, data minimisation, secure transmission and storage of data and manage risks from third parties. If they adopt these practices, the manager will establish a privacy and security system protecting customer data and they will also gain their trust, as well as complying with regulations like GDPR. Achieving business goal of the application as well as keep user privacy will require these efforts.

Read Our Last Blog: https://www.newassignmenthelp.co.uk/leadership-and-the-management-of-project-risk-assignment-sample-53186

Reference List

Journals

• Harwood, M. and Price, R., 2022. Internet and web application security. Jones & Bartlett Learning.
• Bhattacharya, A., Sharma, V., Mishra, R.R. and Sonker, J., 2024. Blockchain Technology in Hospitality: Enhancing Security and Streamlining Operations. In Hotel and Travel Management in the AI Era (pp. 149-178). IGI Global.
• Tan, K.L., Yeap, P.F., Cheong, K.C.K. and Loganathan, S.R., 2025. Crafting an organizational strategy for the new era: a qualitative study of artificial intelligence transformation in a homegrown Singaporean hotel chain. Business Process Management Journal, 31(8), pp.104-123.
• Naidu, S. and Mahmoud, M., 2022, December. Addressing Cloud Computing Security Issues with Solutions. In 2022 International Conference on Computational Science and Computational Intelligence (CSCI) (pp. 1346-1351). IEEE.
• Bhaker, A., 2025. Democratizing Home Ownership through AI-Enabled Financing Tools. Journal of Computer Science and Technology Studies, 7(4), pp.546-555.
• Ranaweera, P., Jurcut, A.D. and Liyanage, M., 2021. Survey on multi-access edge computing security and privacy. IEEE Communications Surveys & Tutorials, 23(2), pp.1078-1124.
• Liveretos, A. and Draganov, I., 2022. Customer Identity and Access Management (CIAM): An overview of the main technology vendors. International Journal of Economics and Management Systems, 7.
• McAmis, R. and Kohno, T., 2023. The writing on the wall and 3d digital twins: personal information in (not so) private real estate. In 32nd USENIX Security Symposium (USENIX Security 23) (pp. 2169-2186).
• Wang, Z. and Liang, J., 2024. Comparative Analysis of Interpretability Techniques for Feature Importance in Credit Risk Assessment. Spectrum of Research, 4(2).
• Nguyen, N.D.K. and Ali, I., 2021. Implementation of cloud customer relationship management in banking sector: Strategies, Benefits and challenges. International Journal of Electronics and Communication Engineering, 15(6), pp.242-247.
• Wang, C., Yuan, Z., Zhou, P., Xu, Z., Li, R. and Wu, D.O., 2023. The security and privacy of mobile-edge computing: An artificial intelligence perspective. IEEE Internet of Things Journal, 10(24), pp.22008-22032.
• Gebraselase, B.G., 2021, March. Blockchain-based information management for network slicing. In 2021 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE) (pp. 555-559). IEEE.
• Jami Pour, M., Irani, H.R. and Yaghobi, A., 2024. Exploring the Challenges of migration towards Software-as-a-Service in Iran: the case study of Cloud-based CRM using a multidimensional perspective. Interdisciplinary Journal of Management Studies (Formerly known as Iranian Journal of Management Studies), 18(1), pp.173-193.
• Sharma, H., 2024. The Evolution of Cybersecurity Challenges and Mitigation Strategies in Cloud Computing Systems. International Journal of Computer Engineering and Technology, 15(4), pp.118-127.
• Mele, C., Di Bernardo, I., Ranieri, A. and Russo Spena, T., 2024. Phygital customer journey: a practice-based approach. Qualitative Market Research: An International Journal, 27(3), pp.388-412.