Notes 3/4/24

 For those who are in 4095 I need a detailed status update of where you are in your build assignments sent to myself and Dr. Bradford.

Mohammed and Ignacio - need to program, program, program

The app engine will have 4 separate functions eventually. You can build each onb separately on different project and then merge together into one.

CompanyInfo

Topics

Graph selection

Conference call loading

Focus on CompanyInfo and Topic first

MOHAMED.OTHMAN@UCONN.EDU

IGNACIO.DELEON@UCONN.EDU

David and Akhil,

YOUNG.NAM@UCONN.EDU

AKHIL.JANNU@UCONN.EDU

You need to get graphs developed We need line graphs of each sentiment classification over 8 periods

Start with the raw code just to draw the graphs> no need to take the data from tables just yet.

After the code is built example 8 periods on bottom graph line

q1 2022, Q2 2022, Q3 2022, Q4 2022, Q1 2023, Q2 2023, Q3 2023, Q4 2023

THE SCORES OF EACH CATEGORY WILL BE PLOTTED against the timeline

Shalini, Arianna and Amid

Need to focus on building both the sentiment details loaded from scores that are currently in spreadsheet

and summary tables

Need to focus on the companies for the sector comparison his should be done for all peers of DHL, WM and UNP with weights 

iN ORDER TO NDO THE GRAPHS WE NEED 8 summary records for each company with 

Yahoo Ticker,

document type,

Yahoo Ticker,

document type,

period,

Average score Macro

Number of times mentioned macro in document

Weighted average score of macro by importance

Average score Sector trend

Number of times mentioned sector trend

Weighted average score sector trend by importance

Average score Financial metric

Number of times mentioned Financial metric

Weighted average financial metric by importance

Average score Regulation

Number of times mentioned regulation

Weighted average regulation score

Total score average

Total score weighted

SHALINI.KUZHIPPAT@UCONN.EDU

ARIANNA.AZIZI@UCONN.EDU

AMID.QAZI@UCONN.EDU

Lastly we need the compute engine ready.

EDWARD.KURYLUK@UCONN.EDU

KENNETH.MAEDA@UCONN.EDU

Teams

 dedeep.singu@uconn.edu

haris.mahmood@uconn.edu

rauni.machado-filho@uconn.edu

charlton.toussaint@uconn.edu

MOHAMED.OTHMAN@UCONN.EDU

IGNACIO.DELEON@UCONN.EDU

AMID.QAZI@UCONN.EDU

SHALINI.KUZHIPPAT@UCONN.EDU

ARIANNA.AZIZI@UCONN.EDU

ananya.jonnakuti@uconn.edu

david.keenan@uconn.edu

YOUNG.NAM@UCONN.EDU

AKHIL.JANNU@UCONN.EDU

EDWARD.KURYLUK@UCONN.EDU

KENNETH.MAEDA@UCONN.EDU

Scoring

The table mentioned below is ‘only for information’ purpose;

it gives summary of ‘number of words’ in each category in the matrix,

and they are not the weights. For example, there are total 18 words in ‘Financial Metric’ category

– of which 7 words are ‘Very Important’ and 6 words are ‘ Important’

and 5 words ‘Less so important’. The table is only for broader understanding.

For our Project purpose:

We have divided all words into three categories namely ‘Very Important’,

‘Important’ and ‘Less so important’ - please refer to attachment and the

color coding. The category classification is to allow the student to ascribe a “weight”

to each of the key words. For example, a key word that has been assign

as ‘Very Important’ should have more weight/importance in the sentiment

score than a key word that is ‘Less so important’.

 

Here’s an illustration of the weight for each of the category.

Very Important -> 2x

Important -> 1x

Less so Important -> 0.5x

Using these weights, the sentiment model score should be more “robust”.

 

We can advise students to be flexible with weights (while maintaining the

priority order) if they find better correlation/relation among peer group.

 

Hope this helps. Let me know if you have any questions.

 

Project Notes 2/18/24

 

To all;

Expecting everyone to report progress and ask questions.

I will be building out more detailed specifications today.

As per prior emails SDP team will be focusing on SSGA deliverables while the 4095 will do the production build.

For build out lets try to get base line programs in place. No need to complete or load all data at this point just get me the code that you are working on explained ASAP.

Key components.

Sentiment Model

This needs to be moved to our Google VM

Haris and Dedeep will need to assist Shalini and Arianna with getting 

sample program running. Does not have to run exactly like it does now. Just get small proof of concept code to work.

Graphs

This is key for the next presentation. Analysts want to see various graphs and comparison. This is how they are vetting the data. We will run graphs from summary tables. Again we need small base line python code. David Keenan and Rauni need to help David Young Nam and Akhil with getting started. Take the data points from the last graphs and pull code out so we can create small samples.

Front end Flask

These should be strait forward and I will build specs. Hopefully team has begun working. Charleton will be helping Mohammed and  Ignacio with the forms. These will run on app engine to load data in tables.

Linux Machine

Charlie and Kenneth will get this machine ready. It is crucial to the project that this gets done ASAP.

Data Loads.

Need to get the Topics data, CompanyInfo, scoring details and scoring totals in tables so programs can work dynamically. Ananya will work with Shalini, Arianna, Amid and Akhil to build and execute loader programs. 

Please provide status and questions as we are half way thru the term and we need to measure our progress.

Topics load

 First task. CRUD - Charlton help Mohammed and Ignacio

Ananya Datastore - Help Shalini and Arianna

Create a CRUD program to store the Datastore table for topics

Leverage the example

https://uconnstamfordslp.blogspot.com/p/assignment-exercise-python-datastore.html

Topics is key

Topics	Classification	Sector	weight
Inflation	Macro	All	Important
Interest rate	Macro	All	Important
Raw Materials inflation	Macro	Industrials	Less so important
Volume	Sector trend	All	Less so important

Create a load program to load the topics spreadsheet into the Datastore table

https://uconn-sa.blogspot.com/p/datastore.html

Leverage the datastore load process 

Note:file is custom keyed

Link to topics spread in share that needs to be loaded.

https://docs.google.com/spreadsheets/d/1N7HwetkXOOaqBDyPWZp1CS0LmOySo0DpzWe3NPpvTpY/edit?usp=sharing

Adjunct Professor John Iacovacci

University of Connecticut, Stamford

John.Iacovacci@uconn.edu