An intelligent sentiment proxy analysis system for estimating price changes in financial markets

Final Year Project

Todo

Results Evaluation
- How many std devs in sentiment for 1 std dev in price
- correlation if near 0 no linear correlation
- Define:
  - LexisNexis articles and their types
  - Gamestop - quote? ordinary company pos->neg->pos return
Discussion and Conclusion
- panel regression
Declaration
max lag should be 5?
changes suggested by Khurshid
dictionary vs ml
ml vs var
state that I chose pearson correlation
more references
reference background in rest of paper
news source improvement: e.g. the importants of non standard media, such as social media
comparison between source types and even specific sources, e.g. newspaper vs social media, twitter vs facebook

Name		Name	Last commit message	Last commit date
Latest commit History 49 Commits
__pycache__		__pycache__
articles		articles
dictionaries		dictionaries
dissertation		dissertation
prices		prices
.env		.env
.gitattributes		.gitattributes
.gitignore		.gitignore
IntelligentAnalysis.py		IntelligentAnalysis.py
README.md		README.md
UserInterface.py		UserInterface.py
VectorAutoregression.py		VectorAutoregression.py
correlation.py		correlation.py
data.xls		data.xls
data.xlsx		data.xlsx
dataCleanUp.py		dataCleanUp.py
dataGraphing.py		dataGraphing.py
dataTabling.py		dataTabling.py
dateUtil.py		dateUtil.py
descriptiveStatistics.py		descriptiveStatistics.py
geckodriver.exe		geckodriver.exe
geckodriver.log		geckodriver.log
modules.txt		modules.txt
priceGatherer.py		priceGatherer.py
priceVsSentiment.jpg		priceVsSentiment.jpg
returnEstimator.py		returnEstimator.py
sentimentExtractor.py		sentimentExtractor.py