Social network analysis 2000 words

Southampton Business School: Postgraduate Module Grade Descriptor

Postgraduate Grade Descriptor for MANG6331 Text Mining and Social Network Analytics

Percentage 0 - 34 35 – 49 50 – 59 60 – 69 70 - 79 80 - 100

Degree Class Fail Compensatable fail*

Pass Merit Distinction Distinction

Collecting unstructured data and conducting exploratory analysis Collecting raw tweets of two different airlines and conducting exploratory data analysis Weighting 20%

No/inadequate evidence of collecting and pre- processing the raw data. No/inadequate evidence of any data analysis.

Evidence of basic but inadequate approaches to collect and/or pre- process the raw data. Mostly descriptive, with minimal data analysis. Argument is basic and poorly constructed.

Collecting and/or pre-processing the raw data is evident but with some confusion. Data analysis is reasonable. Argument is appropriate but with some confusion.

Clear evidence of data pre- processing and exploratory data analysis with minimal omissions/errors. Clear and effective analysis. Argument is structured and is legitimate.

Data pre- processing and exploratory data analysis are appropriate and precise. Comprehensive and precise analysis. Well- structured argument that provides very good clarity. Appropriately use of other sources of information to support arguments.

Data pre- processing and exploratory data analysis are appropriate and precise. Excellent analysis, precise and concise. Exceptionally well- structured argument that provides excellent clarity. Outstanding use of other sources of information to support arguments.

Gaining customer insights: traditional versus social media Evaluate the pros and cons of replacing customer satisfaction survey by mining twitter data Weighting 20%

Not included. Limited and patchy evidence of knowledge and understanding of the pros and cons. Limited evidence of reading. Lacks focus and direction with limited coherent argument.

Sufficient but inconsistent evidence of knowledge and understanding of the pros and cons. Evidence of some use of academic/ business literature. Argument is basic and poorly constructed.

Good knowledge and understanding of the pros and cons. Good use of academic/ business literature to support arguments. Clear and effective argument.

A comprehensive and thorough awareness of the pros and cons. Evidence of comprehensive reading. Well-structured argument that provides very good clarity.

A comprehensive and thorough awareness of the pros and cons. Excellent coverage of relevant literature. Exceptionally well- structured argument that provides excellent clarity.

*Compensatable fail is only possible for compulsory or optional modules, subject to University of Southampton Progression Regulations.

2 Percentage 0 - 34 35 – 49 50 – 59 60 – 69 70 - 79 80 - 100

Degree Class Fail Compensatable fail*

Pass Merit Distinction Distinction

Building advanced analytic models with unstructured data Predicting the polarity of tweet sentiments with two different classification models Weighting 40%

No/inadequate evidence of pre- processing the data. No/inadequate evidence of any data analysis.

Evidence of basic but inadequate approaches to pre- process the data. No/inadequate evidence of any exploratory data analysis. No/inadequate evidence of any classification models. Limited explanation of results.

Evidence of basic data pre- processing/ exploratory data analysis. At least one classification model is evident but with some confusion. No/inadequate evidence of any performance measures. Adequate interpretation of results with some confusion.

Clear evidence of data pre- processing and exploratory data analysis with minimal omissions/errors. Both classification models are evident with minimal omissions/errors. Evidence of some performance measures. Good interpretation of results.

Data pre- processing and exploratory data analysis are appropriate and precise. Both classification models are appropriate and precise. Evidence of appropriate use of performance measures. Excellent interpretation of results.

Data pre- processing and exploratory data analysis are appropriate and precise. Comprehensive and precise classification models are evidence. Evidence of appropriate use of performance measures. Excellent interpretation of results. Exceptionally clear evidence of extensive reading and extend the task significantly beyond core texts.

*Compensatable fail is only possible for compulsory or optional modules, subject to University of Southampton Progression Regulations.

3 Percentage 0 - 34 35 – 49 50 – 59 60 – 69 70 - 79 80 - 100

Degree Class Fail Compensatable fail*

Pass Merit Distinction Distinction

Use of Software Evidence and record of how software was used Weighting 10%

No evidence of using any software

No evidence of using Python or R No Python or R syntax submitted

Submitted Python or R syntax with numerous omissions/errors. Significant deficiencies and/or inconsistencies in comments and/or the structure of the Python or R syntax.

Submitted Python or R syntax with minimal omissions/errors. Some deficiencies and/or inconsistencies in comments and/or the structure of the Python or R syntax.

Mostly well- structured Python or R programme to address the tasks with very good comments.

Excellent comments. Very well-structured and elegant Python or R programme to address the tasks. Professional style.

Structure and Language Well-structured report and good use of languages Weighting 10%

Very poor and often/mostly inarticulate. Mostly incomprehensible.

Significant deficiencies in expression. Inconsistent and poor use of language.

Some deficiencies are apparent. Competent but some inconsistencies apparent.

Mostly well expressed. Clear and appropriate use of language.

Very well expressed. Confident and very good use of language.

Very competent and fluent use of expression. Confident and very good use of language.