7,612
edits
Changes
Jump to navigation
Jump to search
Editing '''5/1''' *Joined patent numbers with inventors -> unipatentinventorcount.sql**Z:univ/InventorCounts.txt*Decade patent counts -> Z:univ/DecadePatentCounts*Calculated Hirschman-Herfindahl index -> E:McNair/University Patents/Inventor Counts and polishing reportSuperstars '''5/5''' Artifacts:#10-Year Ranking All#10-Year Ranking Private#10-Year Ranking Public#Top 10 Movers and Shakers#Graph of Concentration of Innovation#Number of University Patents Over Time#Inventor Concentration#Regression Table *Created Graph of Concentration of Innovation -> E:McNair/University Patents/Inventor Counts and Superstars -> GraphUPS*Created Top Inventors Chart -> University Patents/Report Articles/Inventor Rankings '''5/9''' *Figuring out what's up with the university puller*Developing new 10-Year Rankings*Need to update above two articles*Need to load into STATA and re-run regressions -> University Patents/STATA/univdata.do '''5/10''' *Figured out what was wrong with the data pull and now everything's good!*Updated information on the draft below*Fantastic Artifacts and Where to Find Them**10-Year Ranking All***Chart (University Patents > Report Artifacts > 10-Year University Patent Rankings)**10-Year Ranking Public***Chart (University Patents > Report Artifacts > 10-Year University Patent Rankings)**10-Year Ranking Private***Chart (University Patents > Report Artifacts > 10-Year University Patent Rankings)**Top 10 Movers & Shakers (% change between first 5 years (2006-2010) vs next 5 years (2011-2015) within top 100**Graph of Concentration of Innovation (numbers of patents)***Pie graph (University Patents > Report Artifacts > Concentration of Innovation)***(Excel sheet where this came from is University Patents > Inventor Counts and Concentrations)**Trends Over Time***Total vs. top 10***Graph (University Patents > MoversandTop10)**Inventor Concentration***Chart (University Patents > Report Artifacts > Inventor Rankings)**Reg Table***+Explanation of variables ***(STATA do-file with the most statistically significant regressions is in University Patents > STATA > univdata.do) Issue brief draft (VERY rough): https://docs.google.com/a/rice.edu/document/d/1MayXQEQ_pM0LMeiV39iaNFjtMZYSB-Ah9gCx1rcgW9M/edit?usp=sharing ==Ranking Development Notes==Sources of University Funding R&D [https://www.aaas.org/page/rd-colleges-and-universities]*Roughly 60% comes from the Federal Government*Less than 10% comes from state and local governments*Roughly 20% is from the University itself (endowment?? tuition? not 100% sure)*Roughly 5% is from private industry*Less than 10% Other Federal Government Funding Breakdown [https://www.aaas.org/page/rd-colleges-and-universities]*Roughly 60% of federal funding is from the NIH*Roughly 15% is from the NSF*Roughly 11% from Department of Defense*Roughly 4% from NASA*Roughly 4% from the Department of Energy*Roughly 3% from the USDA*Roughly 1% from the EPA (probably will go to zero with new admin)*Roughly 1% from the Department of Education*Marginal amounts from everywhere else (<1%) NOTES*overall ranking (total number of patents)*weight by faculty (people)*weight by research funding*weight by endowment (maybe)*public vs private*individuals: surname and first initial**ranking by inventors - who are the superstar inventors and where are they (define superstar inventor - idk 20 patents and rank universities by how many superstars are present**Herfindahl-Hirschmann Index
*Patent Counts
===Done===
1. lit review (look at for 1-2 days), look for holistic picture, what has/hasn’t been done? what are the questions people are asking? (DONE)
2. look at patent data, determine how to find universities in patent database: board of regents, universities in patent data (find patterns associated with university assignees) (DONE)
3. get list of patent assignees with countries attached (DONE - Marcela gave us a list of all american assignees)
no edit summary
{{Project|Has project output=Data,Content,Guide|Has sponsor=McNair ProjectsCenter
|Has title=University Patents
|Has owner=Julia Wang, Meghana Pannala,Anne Dayton
|Has keywords=Patent
|Has project status=Active
}}
==Research Paper==
Since university system file for patents out of a central office, its mot possible to determine where the research was conducted. This presents problems when ranking universities by patent production.
In order to publish this paper as a Baker Institute Research Report, we will need to do more to resolve this issue beyond weighting for the University of California.
==Carnegie Classifications==
We propose using public data from the Carnegie Classifications of Institutes of Higher Education to resolve the university systems issue.
TASKS:
1) Separate schools affiliated with universities/university systems in our rankings from all others included in Carnegie data.
2)Group these by system.
Determine the number of schools in each system
3) Determine their classification.
1 if Doctoral 0 if not (In Carnegie 2015 Basic Classification [BASIC2015] 15, 16, 17)
1 if 4-year medical school 0 if not (BASIC2015 25) Check (BASIC2015 26)
1 if 4-year engineering 0 if not (BASIC2015 27) Check (BASIC2015 28)
4) Determine the total count of the categories in number 3 for each university system.
5) Use total count to weight regressions
6) Create ranking tables that rank university systems and universities by dividing number of patents by number of doctoral, med-school, 4-year engineering.
==Project Overview==
Goal: list of all universities and # of patents associated with each university and patent licensing activity
AUTM?
==Timeline to Deliverable& Documentation==
Meghana: focus on grants
Julia: focus on patent counts
Working in E:McNair/University Patents '''3/20-3/24 '''
Learn SQL, clean data
'''3/27-3/31 '''
Counts of data, name matching
Patent Counts*Ran the matcher in the server, only matched ~40,000 entries*Working with Jeemin to develop matcher '''4/3-4/7'''
Counts of data, name matching
Patent Counts*Jeemin is a savior, matched and counted 128,000 entries (Jeemin_matcher_matched.txt), manually went through the 6,000 unmatched (Jeemin_matcher_unmatched.txt), Check [[University Patent Matching]]*Ranking for all time is in Patent Counting ( >Total (All Time) )*Need to talk with Meghana about grant data*Need to find variables for regression, develop ranking for last decade and last year*Also need to account for school size (faculty? students? research funding?) '''4/10-4/14'''
Develop ranking
Patent Counts*Finish all of above*Start drafting*Load data onto database -> E:McNair/University Patents/unidatajoin.sql in univ (database) '''4/17-4/21''' *Combined Patent Counts, R&D Expenditure, NSF Grant Data*STATA regressions in E:McNair/University Patents/STATA**for historical data '''4/24''' From Ed:*Copied inventors table out of PatentDB database on RDP*Script in E:McNair/PatentData, MoveInventors.sql*Copy of data in Z:AllPatents and ran import into patent (database)For Marcela:*Need table of all patents associated with matches, join patent numbers with inventors *Group by last name and first initial (=1 inventor)
==Key Words==
**Quality of TTO
**Publications (specifically based on research)?
**Amount received in grants/Amount of grants - we have numbers on amount of NIH grants and NSF grants
***NIH/NSF/STTR
**Control for school size/endowment
*Grant Data
*Research Funding
*Star Scientists
*Regression with variables from Carnegie Classification (10-15 variables)
*Licensing (?)
===What We Have===
*Lit Review
*List of all post-secondary institutions (University Accreditation 3.2016_trim) - U.S. Department of Education
*Classification of universities (Carnegie Classification 2015) - Carnegie Classification
*Jeemin has found all the correct and incorrect spellings of assignees with "University"
*Julia working on "Institute"
*We have a list of Universities that have patents from NIH grants, which could be a starting point for our ranking constituents
==Lit Reviews==
