Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/gplacide/pagerank
PageRank Algorithm
https://github.com/gplacide/pagerank
gensim-library gensim-word2vec pagerank word2vec
Last synced: about 2 months ago
JSON representation
PageRank Algorithm
- Host: GitHub
- URL: https://github.com/gplacide/pagerank
- Owner: gPlacide
- Created: 2020-09-10T21:50:52.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2020-12-06T20:39:06.000Z (about 4 years ago)
- Last Synced: 2023-10-20T01:30:29.063Z (about 1 year ago)
- Topics: gensim-library, gensim-word2vec, pagerank, word2vec
- Language: Python
- Homepage:
- Size: 9.67 MB
- Stars: 9
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Pagerank Algorithm
In this project, I create a simple search engine for the Lawfare blog (https://www.lawfareblog.com/). I later updated this project by including a word2vec library ```gensim``` to improve the search engine where webpages containing related words to the searched key term are also returned. For the examples below I used ```'glove-wiki-gigaword-300'``` model in ```gensim```.
# Step 1: The Power Method
I implimented ```WebGraph.power_method``` function found in ```pagerank.py``` to compute the pagerank vector.
Initially, I run the power method on a smaller graph ```small.csv.gz``` that was in the paper *Deeper Inside Pagerank*. Below was my output:
```
python3 pagerank.py --data=./small.csv.gz
INFO:root:rank=0 pagerank=2.1634e+00 url=4
INFO:root:rank=1 pagerank=1.6664e+00 url=6
INFO:root:rank=2 pagerank=1.2402e+00 url=5
INFO:root:rank=3 pagerank=4.5712e-01 url=2
INFO:root:rank=4 pagerank=3.5620e-01 url=3
INFO:root:rank=5 pagerank=3.2078e-01 url=1
````
After my Power method was working correctly, I moved on to the Lawfare blog and started using the power method to return urls matching the entered query sorted according to their pagerank.
```--search_query``` == **corona**
```
python3.8 pagerank.py --data=./lawfareblog.csv.gz --search_query='corona'
INFO:gensim.models.keyedvectors:precomputing L2-norms of word weight vectors
INFO:root:rank=0 pagerank=1.0038e-03 url=www.lawfareblog.com/lawfare-podcast-united-nations-and-coronavirus-crisis
INFO:root:rank=1 pagerank=8.9224e-04 url=www.lawfareblog.com/house-oversight-committee-holds-day-two-hearing-government-coronavirus-response
INFO:root:rank=2 pagerank=7.0390e-04 url=www.lawfareblog.com/britains-coronavirus-response
INFO:root:rank=3 pagerank=6.9153e-04 url=www.lawfareblog.com/prosecuting-purposeful-coronavirus-exposure-terrorism
INFO:root:rank=4 pagerank=6.7041e-04 url=www.lawfareblog.com/israeli-emergency-regulations-location-tracking-coronavirus-carriers
INFO:root:rank=5 pagerank=6.6256e-04 url=www.lawfareblog.com/why-congress-conducting-business-usual-face-coronavirus
INFO:root:rank=6 pagerank=6.5046e-04 url=www.lawfareblog.com/congressional-homeland-security-committees-seek-ways-support-state-federal-responses-coronavirus
INFO:root:rank=7 pagerank=6.3620e-04 url=www.lawfareblog.com/paper-hearing-experts-debate-digital-contact-tracing-and-coronavirus-privacy-concerns
INFO:root:rank=8 pagerank=6.1248e-04 url=www.lawfareblog.com/house-subcommittee-voices-concerns-over-us-management-coronavirus
INFO:root:rank=9 pagerank=6.0187e-04 url=www.lawfareblog.com/livestream-house-oversight-committee-holds-hearing-government-coronavirus-response
```
```--search-query``` == **trump**
```
python3.8 pagerank.py --data=./lawfareblog.csv.gz --search_query='trump'
INFO:gensim.models.keyedvectors:precomputing L2-norms of word weight vectors
INFO:root:rank=0 pagerank=5.7826e-03 url=www.lawfareblog.com/trump-asks-supreme-court-stay-congressional-subpeona-tax-returns
INFO:root:rank=1 pagerank=5.2338e-03 url=www.lawfareblog.com/document-trump-revokes-obama-executive-order-counterterrorism-strike-casualty-reporting
INFO:root:rank=2 pagerank=5.1297e-03 url=www.lawfareblog.com/trump-administrations-worrying-new-policy-israeli-settlements
INFO:root:rank=3 pagerank=4.6599e-03 url=www.lawfareblog.com/dc-circuit-overrules-district-courts-due-process-ruling-qasim-v-trump
INFO:root:rank=4 pagerank=4.5934e-03 url=www.lawfareblog.com/donald-trump-and-politically-weaponized-executive-branch
INFO:root:rank=5 pagerank=4.3071e-03 url=www.lawfareblog.com/how-trumps-approach-middle-east-ignores-past-future-and-human-condition
INFO:root:rank=6 pagerank=4.0935e-03 url=www.lawfareblog.com/why-trump-cant-buy-greenland
INFO:root:rank=7 pagerank=3.7591e-03 url=www.lawfareblog.com/oral-argument-summary-qassim-v-trump
INFO:root:rank=8 pagerank=3.4509e-03 url=www.lawfareblog.com/dc-circuit-court-denies-trump-rehearing-mazars-case
INFO:root:rank=9 pagerank=3.4484e-03 url=www.lawfareblog.com/second-circuit-rules-mazars-must-hand-over-trump-tax-returns-new-york-prosecutors
```
```--search-query``` == **iran**
```
python3.8 pagerank.py --data=./lawfareblog.csv.gz --search_query='iran'
INFO:gensim.models.keyedvectors:precomputing L2-norms of word weight vectors
INFO:root:rank=0 pagerank=4.5746e-03 url=www.lawfareblog.com/praise-presidents-iran-tweets
INFO:root:rank=1 pagerank=4.4174e-03 url=www.lawfareblog.com/how-us-iran-tensions-could-disrupt-iraqs-fragile-peace
INFO:root:rank=2 pagerank=2.6928e-03 url=www.lawfareblog.com/cyber-command-operational-update-clarifying-june-2019-iran-operation
INFO:root:rank=3 pagerank=1.9391e-03 url=www.lawfareblog.com/aborted-iran-strike-fine-line-between-necessity-and-revenge
INFO:root:rank=4 pagerank=1.8263e-03 url=www.lawfareblog.com/its-not-only-iraq-and-syria
INFO:root:rank=5 pagerank=1.5503e-03 url=www.lawfareblog.com/readings-post-syria-extracted-onion
INFO:root:rank=6 pagerank=1.5452e-03 url=www.lawfareblog.com/parsing-state-departments-letter-use-force-against-iran
INFO:root:rank=7 pagerank=1.5357e-03 url=www.lawfareblog.com/iranian-hostage-crisis-and-its-effect-american-politics
INFO:root:rank=8 pagerank=1.5258e-03 url=www.lawfareblog.com/announcing-united-states-and-use-force-against-iran-new-lawfare-e-book
INFO:root:rank=9 pagerank=1.4962e-03 url=www.lawfareblog.com/congress-us-policy-toward-syria-and-turkey-overview-recent-hearings
```
Next, I wanted to figure out what were the most important articles on the blog. However, pages on the blog got higher pageranks because they all have links to the main blog url.
```
python3 pagerank.py --data=./lawfareblog.csv.gz
INFO:root:rank=0 pagerank=8.4156e+00 url=www.lawfareblog.com/litigation-documents-related-appointment-matthew-whitaker-acting-attorney-general
INFO:root:rank=1 pagerank=8.4156e+00 url=www.lawfareblog.com/lawfare-job-board
INFO:root:rank=2 pagerank=8.4156e+00 url=www.lawfareblog.com/documents-related-mueller-investigation
INFO:root:rank=3 pagerank=8.4156e+00 url=www.lawfareblog.com/litigation-documents-resources-related-travel-ban
INFO:root:rank=4 pagerank=8.4156e+00 url=www.lawfareblog.com/subscribe-lawfare
INFO:root:rank=5 pagerank=8.4156e+00 url=www.lawfareblog.com/masthead
INFO:root:rank=6 pagerank=8.4156e+00 url=www.lawfareblog.com/topics
INFO:root:rank=7 pagerank=8.4156e+00 url=www.lawfareblog.com/our-comments-policy
INFO:root:rank=8 pagerank=8.4156e+00 url=www.lawfareblog.com/upcoming-events
INFO:root:rank=9 pagerank=8.4156e+00 url=www.lawfareblog.com/about-lawfare-brief-history-term-and-site
```
You can see that the power method returned a good number of links that are pages instead of articles that I am more interested in. Hence, I used ```--filter-ratio``` to remove urls that have more links than the specified fraction.
```
python3 pagerank.py --data=./lawfareblog.csv.gz --filter_ratio=0.2
INFO:root:rank=0 pagerank=4.6091e+00 url=www.lawfareblog.com/trump-asks-supreme-court-stay-congressional-subpeona-tax-returns
INFO:root:rank=1 pagerank=2.9867e+00 url=www.lawfareblog.com/livestream-nov-21-impeachment-hearings-0
INFO:root:rank=2 pagerank=2.9669e+00 url=www.lawfareblog.com/opening-statement-david-holmes
INFO:root:rank=3 pagerank=2.0173e+00 url=www.lawfareblog.com/senate-examines-threats-homeland
INFO:root:rank=4 pagerank=1.8769e+00 url=www.lawfareblog.com/what-make-first-day-impeachment-hearings
INFO:root:rank=5 pagerank=1.8762e+00 url=www.lawfareblog.com/livestream-house-armed-services-committee-hearing-f-35-program
INFO:root:rank=6 pagerank=1.8693e+00 url=www.lawfareblog.com/whats-house-resolution-impeachment
INFO:root:rank=7 pagerank=1.7655e+00 url=www.lawfareblog.com/congress-us-policy-toward-syria-and-turkey-overview-recent-hearings
INFO:root:rank=8 pagerank=1.6807e+00 url=www.lawfareblog.com/summary-david-holmess-deposition-testimony
```
The paper, *Deeper Inside Pagerank* suggests that alpha of .85 is preferable. Let's rerun the above chunk with a different alpha value and see how the pagerank will be affected.
```
python3 pagerank.py --data=./lawfareblog.csv.gz --verbose --filter_ratio=0.2 --alpha=0.99999
INFO:root:rank=0 pagerank=5.2385e+01 url=www.lawfareblog.com/lawfare-live-covid-19-speech-and-surveillance
INFO:root:rank=1 pagerank=5.2385e+01 url=www.lawfareblog.com/covid-19-speech-and-surveillance-response
INFO:root:rank=2 pagerank=7.9438e+00 url=www.lawfareblog.com/cost-using-zero-days
INFO:root:rank=3 pagerank=2.3700e+00 url=www.lawfareblog.com/lawfare-podcast-former-congressman-brian-baird-and-daniel-schuman-how-congress-can-continue-function
INFO:root:rank=4 pagerank=1.5529e+00 url=www.lawfareblog.com/events
INFO:root:rank=5 pagerank=1.1867e+00 url=www.lawfareblog.com/water-wars-increased-us-focus-indo-pacific
INFO:root:rank=6 pagerank=1.1867e+00 url=www.lawfareblog.com/water-wars-drill-maybe-drill
INFO:root:rank=7 pagerank=1.1867e+00 url=www.lawfareblog.com/water-wars-disjointed-operations-south-china-sea
INFO:root:rank=8 pagerank=1.1867e+00 url=www.lawfareblog.com/water-wars-us-china-divide-shangri-la
INFO:root:rank=9 pagerank=1.1867e+00 url=www.lawfareblog.com/water-wars-sinking-feeling-philippine-china-relations
```
It's clear that increasing alpha from .85 to .99999 changed the pagerank and it would be up to the one who is writing the algorithm to decide which pagerank makes more sense than the other.
# Step 2: The personalization vector
I started by implementing ```WebGraph.make_personalization_vector``` function. This function enables ```--personalization_vector_query``` argument which makes changes to the personalization vector.
Earlier when we used ```--search_query``` we got
```
python3.8 pagerank.py --data=./lawfareblog.csv.gz --search_query='corona'
INFO:gensim.models.keyedvectors:precomputing L2-norms of word weight vectors
INFO:root:rank=0 pagerank=1.0038e-03 url=www.lawfareblog.com/lawfare-podcast-united-nations-and-coronavirus-crisis
INFO:root:rank=1 pagerank=8.9224e-04 url=www.lawfareblog.com/house-oversight-committee-holds-day-two-hearing-government-coronavirus-response
INFO:root:rank=2 pagerank=7.0390e-04 url=www.lawfareblog.com/britains-coronavirus-response
INFO:root:rank=3 pagerank=6.9153e-04 url=www.lawfareblog.com/prosecuting-purposeful-coronavirus-exposure-terrorism
INFO:root:rank=4 pagerank=6.7041e-04 url=www.lawfareblog.com/israeli-emergency-regulations-location-tracking-coronavirus-carriers
INFO:root:rank=5 pagerank=6.6256e-04 url=www.lawfareblog.com/why-congress-conducting-business-usual-face-coronavirus
INFO:root:rank=6 pagerank=6.5046e-04 url=www.lawfareblog.com/congressional-homeland-security-committees-seek-ways-support-state-federal-responses-coronavirus
INFO:root:rank=7 pagerank=6.3620e-04 url=www.lawfareblog.com/paper-hearing-experts-debate-digital-contact-tracing-and-coronavirus-privacy-concerns
INFO:root:rank=8 pagerank=6.1248e-04 url=www.lawfareblog.com/house-subcommittee-voices-concerns-over-us-management-coronavirus
INFO:root:rank=9 pagerank=6.0187e-04 url=www.lawfareblog.com/livestream-house-oversight-committee-holds-hearing-government-coronavirus-response
```
but after using ```--personalization_vector_query``` we get
```
python3.8 pagerank.py --data=./lawfareblog.csv.gz --filter_ratio=0.2 --personalization_vector_query='corona'--search_query='corona'
INFO:gensim.models.keyedvectors:precomputing L2-norms of word weight vectors
INFO:root:rank=0 pagerank=1.2209e-01 url=www.lawfareblog.com/brexit-not-immune-coronavirus
INFO:root:rank=1 pagerank=1.2209e-01 url=www.lawfareblog.com/rational-security-my-corona-edition
INFO:root:rank=2 pagerank=9.1920e-02 url=www.lawfareblog.com/prosecuting-purposeful-coronavirus-exposure-terrorism
INFO:root:rank=3 pagerank=9.1920e-02 url=www.lawfareblog.com/britains-coronavirus-response
INFO:root:rank=4 pagerank=7.7770e-02 url=www.lawfareblog.com/lawfare-podcast-united-nations-and-coronavirus-crisis
INFO:root:rank=5 pagerank=7.2888e-02 url=www.lawfareblog.com/house-oversight-committee-holds-day-two-hearing-government-coronavirus-response
INFO:root:rank=6 pagerank=6.9239e-02 url=www.lawfareblog.com/livestream-house-oversight-committee-holds-hearing-government-coronavirus-response
INFO:root:rank=7 pagerank=6.8777e-02 url=www.lawfareblog.com/paper-hearing-experts-debate-digital-contact-tracing-and-coronavirus-privacy-concerns
INFO:root:rank=8 pagerank=6.5438e-02 url=www.lawfareblog.com/israeli-emergency-regulations-location-tracking-coronavirus-carriers
INFO:root:rank=9 pagerank=4.7884e-02 url=www.lawfareblog.com/china-responds-coronavirus-iron-grip-information-flow
```
In the latter example, a webpage is important only if other coronavirus webpages think it's important while the first example treats a webpage as important when any other webpage thinks that it's important.
Lastly, let's look at webpages that are related to ```corona``` or ```iran``` but that do not directly mention those two words. These webpages are relevant to the topics but because our ```--search_query``` had to include corona or iran we did not get them in our pagerank.
```
python3 pagerank.py --data=./lawfareblog.csv.gz --filter_ratio=0.2 --personalization_vector_query='corona' --search_query='-corona'
INFO:root:rank=0 pagerank=8.8870e-01 url=www.lawfareblog.com/covid-19-speech-and-surveillance-response
INFO:root:rank=1 pagerank=8.8867e-01 url=www.lawfareblog.com/lawfare-live-covid-19-speech-and-surveillance
INFO:root:rank=2 pagerank=1.8256e-01 url=www.lawfareblog.com/chinatalk-how-party-takes-its-propaganda-global
INFO:root:rank=3 pagerank=1.0729e-01 url=www.lawfareblog.com/trump-cant-reopen-country-over-state-objections
INFO:root:rank=4 pagerank=9.4298e-02 url=www.lawfareblog.com/lawfare-podcast-mom-and-dad-talk-clinical-trials-pandemic
INFO:root:rank=5 pagerank=7.9633e-02 url=www.lawfareblog.com/fault-lines-foreign-policy-quarantined
INFO:root:rank=6 pagerank=7.5307e-02 url=www.lawfareblog.com/limits-world-health-organization
INFO:root:rank=7 pagerank=6.8115e-02 url=www.lawfareblog.com/chinatalk-dispatches-shanghai-beijing-and-hong-kong
INFO:root:rank=8 pagerank=6.4847e-02 url=www.lawfareblog.com/us-moves-dismiss-case-against-company-linked-ira-troll-farm
INFO:root:rank=9 pagerank=6.4847e-02 url=www.lawfareblog.com/livestream-house-foreign-affairs-committee-holds-hearing-crisis-idlib
```
```
python3 pagerank.py --data=./lawfareblog.csv.gz --filter_ratio=0.2 --personalization_vector_query='iran' --search_query='-iran'
INFO:root:rank=0 pagerank=4.5063e-01 url=www.lawfareblog.com/omphalos
INFO:root:rank=1 pagerank=2.5712e-01 url=www.lawfareblog.com/cancellation-algerias-elections-opportunity-democratization
INFO:root:rank=2 pagerank=2.5394e-01 url=www.lawfareblog.com/yemen-houthi-strategy-has-promise-and-risk
INFO:root:rank=3 pagerank=2.5307e-01 url=www.lawfareblog.com/haftar-attacking-tripoli-us-needs-re-engage-libya
INFO:root:rank=4 pagerank=2.5307e-01 url=www.lawfareblog.com/how-trumps-approach-middle-east-ignores-past-future-and-human-condition
INFO:root:rank=5 pagerank=2.0710e-01 url=www.lawfareblog.com/trump-asks-supreme-court-stay-congressional-subpeona-tax-returns
INFO:root:rank=6 pagerank=1.9135e-01 url=www.lawfareblog.com/blurred-distinction-between-armed-conflict-and-civil-unrest-recent-events-gaza
INFO:root:rank=7 pagerank=1.8959e-01 url=www.lawfareblog.com/document-sen-tim-kaine-presses-pentagon-legal-definition-collective-self-defense
INFO:root:rank=8 pagerank=1.8959e-01 url=www.lawfareblog.com/document-july-2018-nato-summit-communique
INFO:root:rank=9 pagerank=1.8942e-01 url=www.lawfareblog.com/al-kibar-strike-what-difference-26-years-make
```