ANLY482 Current Practicums: Current Practicums

About

Current Practicums

Past Practicums

Grading & Deliverables

Downloads & FAQ

Current Run

Team	About the Project	Student Member(s)	Project Supervisor	Sponsor
Terrorism : Unraveling Hidden Patterns	The purpose of the practicum is to acquaint students with quantitative studies of terrorism. The dataset includes the merger of two kinds of datasets. The first is a dataset of terrorist organizations and their attributes (ideology, size, age, funding, etc). The second is a dataset of events or terrorist attacks and their attributes (date, target, weapon used, people killed, etc) from 1967-2011. The goal is to 1) find a stable pattern, 2) display it visually, and 3) attempt to explain it or at least identify some factors that correlate with it. Those should be theoretically informed. You can also display the way the factors are correlated with explaining the pattern. Types of analysis: Theme: At the big-picture level, the project should focus on two aspects of terrorism: tactics and targets of terrorist organizations. For example you may consider how a particular tactic such as the use of car bombs has spread across the globe. Alternatively you may examine how there has been a shift in targets over time. Here are some ideas that express this theme: Idea 1: One particular area that is relevant to Singapore is what I call Proxy Terrorism, which is terrorism aimed at a third-party. For most terrorist acts the targets are coterminous with the countries that are being attacked. The target in country A is meant to punish country A. However, some terrorist acts seek to attack targets in country A to punish country B. Examples of such attacks include tourist areas and embassies. In Singapore’s history the single successful terrorist attack – the Laju Incident – was a proxy attack. Similarly plots by Jemaah Islamiyah, that were successfully foiled, were against foreign embassies in Singapore. Neighborhood groups in Indonesia and Malaysia also engaged in proxy terrorism against tourist resorts in those countries or other proxy targets. To illustrate with the above example, you may find that proxy terrorist attacks have declined precipitously or are clustered in a particular geographic area. You might display a graph over time and you may identify some factors or some conditions that make proxy terrorist attacks more likely. Idea 2: Most terrorist organizations attack both civilians and military targets. You might find that a given terrorist organization’s civilian:military ratio tends to increase over the lifespan of the terrorist organization. You might illustrate this by displaying a series of graphs over time. It may be that terrorist organizations that are ethnonationalist are more likely to have this ratio increase, while terrorist organizations that are religious tend to have this ratio decrease. We will soon share a document with more ideas to help you. Suggested Platforms: SAS Visual Analytics suite; tableau or D3.js for developing custom visualizations Libraries to explore: SAS VA, D3.js, C3.js Other recommendations: No prior knowledge of any courses is assumed while designing this project but prior knowledge of Visual Analytics will be good for taking up advanced analysis during project related tasks. Recommended Team Composition: Students are free to come up with their own teams but forming a team with diverse backgrounds and skill-sets is highly recommended. More information on the dataset will be shared soon.		Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics	Prof. Michael GENKIN Assistant Professor of Sociology, SOSS
Improved Decisions for Ocean Freights	The Green Transformation Lab (GTL) is a joint initiative by SMU and DHL aimed at accelerating the evolution of sustainable logistics across Asia Pacific. Leveraging SMU’s multi-faculty academic excellence and DHL’s sustainability services, expertise and capability in supply chains, the Green Transformation Lab is focused on creating solutions that help companies transform their supply chains, becoming greener, more resource efficient and sustainable. As part of on-going effort to improve fill-rate of ocean freight, GTL needs a dashboard to visualize and allow decision-makers to select the most appropriate mode of shipment. There are two modes of ocean freight shipments: FCL (Full-container-load) and LCL (Less-than-container-load). More information about general things that GTL does can be found at GTL official website Types of analysis: The aim of this analytics project is to perform data analytics and build a dashboard using a visualization tool such as Tableau. The features on the visualization tool includes (but not limited to): a. Ocean freight profiling for the selected customer b. Visualization of major trade-lanes (pair of source and destination of shipments) c. Simple what-if analysis of a different mode of shipment is selected (e.g., customer choose LCL instead of FCL) on key defined KPIs such as cost, CO2 emissions Suggested Platforms: Project Sponsor would like the team to use Tableau for developing visualizations ; Other recommendations: No prior knowledge of any courses is assumed while designing this project but prior knowledge Visual Analytics will be good for taking up advanced analysis during project related tasks. Recommended Team Composition: Students are free to come up with their own teams but forming a team with diverse backgrounds and skill-sets is highly recommended. More information on the dataset will be shared soon.	Terence CHU Tailun PHANG Ming Min NG Zhen Jie	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics	Prof. TAN Kar Way Assistant Professor of Information Systems (Practice)
Recommendations Matter to Us! - Team BEK	Recommendations and opinions from others are a part of our daily lives. From eating out in a restaurants to buying commodities online we yearn to know what others have to say about them. We want to see what others have experienced from a purchase partly because it involves spending our hard earned money but largely because of the plethora of information on available options out there. Seeking recommendations is an attempt towards staying away from bad experiences and maximizing once sense of satisfaction from a purchase. This project shares people's opinion and their ratings on Yelp about businesses operating in various cites from U.K., Germany, Canada to Unites States. It will give you an opportunity to apply exploratory and predictive analytics techniques such as n-gram analysis, topic identification, sentiment analysis, ratings prediction, item-item similarities to design a flow that can help users across the world find the place of their choice faster with improved precision. Stress will be given on how you design a general purpose model using restaurant related data which can be applied to any other practice in hospitality domain such as hospitals, hotels equally. Types of analysis: The teams can analyse the data to unravel many aspects such as: 1. What seasonal trends and patterns can be detected in data. Are there places being reviewed only on certain occasions or time periods around the year. Are their cities where people tend to eat out more then people from other cities. Try to predict what type of restaurants or bars will be reviewed more based on an upcoming event or festival. 2. Clustering similar businesses, Item-Item to user-user similarities to develop and refine recommendations. 3. Sentiment analysis of English text, use of n-gram analysis to filter out prominent phrases and tips from the reviews, identification of topics of discussions. Identifying sarcasm and sarcastic reviewers. Identifying what are the main complainants, suggestions and wishes of reviewers. 4. Apply above mentioned analysis techniques to warn businesses when their overall image starts to go down, highlights new ideas for them to increase sales. By using a dashboard to display findings. Suggested Platforms: SAS Visual Analytics suite; SAS EM, python or R for text analytics; tableau or D3.js for developing visualizations Libraries to explore: SAS EM and SAS VA sentiment analysis package, Natural language tool kit & libraries (both Python 2.7 and R), scikit-learn (Machine Learning in Python), D3.js, C3.js Other recommendations: No prior knowledge of any courses is assumed while designing this project but prior knowledge of Social and Contextual Analytics, Visual Analytics will be good for taking up advanced analysis during project related tasks. Recommended Team Composition: Students are free to come up with their own teams but forming a team with diverse backgrounds and skill-sets is highly recommended. More information on the dataset in this File:ReferenceDocumentYelp.docx	Keith TAN Xiang Wei YAO Min Gee ZHENG Boyang	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics
Recommendations Matter to Us! - Group 2	Recommendations and opinions from others are a part of our daily lives. From eating out in a restaurants to buying commodities online we yearn to know what others have to say about them. We want to see what others have experienced from a purchase partly because it involves spending our hard earned money but largely because of the plethora of information on available options out there. Seeking recommendations is an attempt towards staying away from bad experiences and maximizing once sense of satisfaction from a purchase. This project shares people's opinion and their ratings on Yelp about businesses operating in various cites from U.K., Germany, Canada to Unites States. It will give you an opportunity to apply exploratory and predictive analytics techniques such as n-gram analysis, topic identification, sentiment analysis, ratings prediction, item-item similarities to design a flow that can help users across the world find the place of their choice faster with improved precision. Stress will be given on how you design a general purpose model using restaurant related data which can be applied to any other practice in hospitality domain such as hospitals, hotels equally. Types of analysis: The teams can analyse the data to unravel many aspects such as: 1. What seasonal trends and patterns can be detected in data. Are there places being reviewed only on certain occasions or time periods around the year. Are their cities where people tend to eat out more then people from other cities. Try to predict what type of restaurants or bars will be reviewed more based on an upcoming event or festival. 2. Clustering similar businesses, Item-Item to user-user similarities to develop and refine recommendations. 3. Sentiment analysis of English text, use of n-gram analysis to filter out prominent phrases and tips from the reviews, identification of topics of discussions. Identifying sarcasm and sarcastic reviewers. Identifying what are the main complainants, suggestions and wishes of reviewers. 4. Apply above mentioned analysis techniques to warn businesses when their overall image starts to go down, highlights new ideas for them to increase sales. By using a dashboard to display findings. Suggested Platforms: SAS Visual Analytics suite; SAS EM, python or R for text analytics; tableau or D3.js for developing visualizations Libraries to explore: SAS EM and SAS VA sentiment analysis package, Natural language tool kit & libraries (both Python 2.7 and R), scikit-learn (Machine Learning in Python), D3.js, C3.js Other recommendations: No prior knowledge of any courses is assumed while designing this project but prior knowledge of Social and Contextual Analytics, Visual Analytics will be good for taking up advanced analysis during project related tasks. Recommended Team Composition: Students are free to come up with their own teams but forming a team with diverse backgrounds and skill-sets is highly recommended. More information on the dataset in this File:ReferenceDocumentYelp.docx	Li Xiang Rhea Chandra Piyush Pritam Sahoo Malvania Smeet Saunil	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics
Recommendations Matter to Us !! - Group 3	Recommendations and opinions from others are a part of our daily lives. From eating out in a restaurants to buying commodities online we yearn to know what others have to say about them. We want to see what others have experienced from a purchase partly because it involves spending our hard earned money but largely because of the plethora of information on available options out there. Seeking recommendations is an attempt towards staying away from bad experiences and maximizing once sense of satisfaction from a purchase. This project shares people's opinion and their ratings on Yelp about businesses operating in various cites from U.K., Germany, Canada to Unites States. It will give you an opportunity to apply exploratory and predictive analytics techniques such as n-gram analysis, topic identification, sentiment analysis, ratings prediction, item-item similarities to design a flow that can help users across the world find the place of their choice faster with improved precision. Stress will be given on how you design a general purpose model using restaurant related data which can be applied to any other practice in hospitality domain such as hospitals, hotels equally. Types of analysis: The teams can analyse the data to unravel many aspects such as: 1. What seasonal trends and patterns can be detected in data. Are there places being reviewed only on certain occasions or time periods around the year. Are their cities where people tend to eat out more then people from other cities. Try to predict what type of restaurants or bars will be reviewed more based on an upcoming event or festival. 2. Clustering similar businesses, Item-Item to user-user similarities to develop and refine recommendations. 3. Sentiment analysis of English text, use of n-gram analysis to filter out prominent phrases and tips from the reviews, identification of topics of discussions. Identifying sarcasm and sarcastic reviewers. Identifying what are the main complainants, suggestions and wishes of reviewers. 4. Apply above mentioned analysis techniques to warn businesses when their overall image starts to go down, highlights new ideas for them to increase sales. By using a dashboard to display findings. Suggested Platforms: SAS Visual Analytics suite; SAS EM, python or R for text analytics; tableau or D3.js for developing visualizations Libraries to explore: SAS EM and SAS VA sentiment analysis package, Natural language tool kit & libraries (both Python 2.7 and R), scikit-learn (Machine Learning in Python), D3.js, C3.js Other recommendations: No prior knowledge of any courses is assumed while designing this project but prior knowledge of Social and Contextual Analytics, Visual Analytics will be good for taking up advanced analysis during project related tasks. Recommended Team Composition: Students are free to come up with their own teams but forming a team with diverse backgrounds and skill-sets is highly recommended. More information on the dataset in this File:ReferenceDocumentYelp.docx	Ng Hui Ying Aldred Lau Wen Yang Michelle Teo Sok Lee Tan Yi Hao	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics	Prof. Seema Chokshi Lecturer of Information Systems, Programme Head, SMU Undergraduate Second Major in Analytics