MétaCan
Menu
Back to cohort
Record W4393650178 · doi:10.5281/zenodo.3594642

DISCERN 2: Duke Innovation & SCientific Enterprises Research Network

2024· dataset· en· W4393650178 on OpenAlexaff
Ashish Arora, Sharon Belenzon

Bibliographic record

VenueZenodo (CERN European Organization for Nuclear Research) · 2024
Typedataset
Languageen
FieldBusiness, Management and Accounting
TopicBig Data and Business Intelligence
Canadian institutionsUniversity of Toronto
Fundersnot available
KeywordsBusinessIndustrial organizationKnowledge managementComputer science

Abstract

fetched live from OpenAlex

The DISCERN dataset was developed to support academic research on corporate innovation by linking data on U.S. publicly listed firms from Standard & Poor’s Compustat database to their patents and scientific publications. A key feature of DISCERN is its comprehensive coverage of firms’ subsidiaries and their ownership changes over time, which is crucial for accurately mapping corporate innovation. Patents and publications may be assigned to various legal entities within a firm’s organizational structure. Subsidiaries may change ownership in M&A events. By accounting for these ownership linkages over time, DISCERN enables researchers to construct more precise measures of firms’ knowledge production and examine the factors influencing their R&D investment decisions. Version 2.0 incorporates several key improvements over the previous version of DISCERN. First, we shift to using the PatentsView database as the main source of patent data and OpenAlex as the main source of scientific publication data. PatentsView is publicly available and continuously maintained directly by the United States Patents & Trademarks Office (USPTO). OpenAlex is currently the only open data source of scientific publication metadata. Using freely available data sources allows us to share both the patent and the publication datasets openly. This enhances data access, which was previously limited due to the use of propriety data. Second, the updated dataset now covers the period from 1980 to 2021, providing an additional six years of data. Third, we transition to using Securities and Exchange Commission (SEC) filings as the primary source of subsidiary data, allowing us to trace ownership linkages further back to the mid-1990s and ensuring a higher degree of reliability compared to the Orbis data used in the original version, which was less reliable and had comprehensive coverage only from 2008. Finally, by transitioning to PatentsView and additional data sourced from the USPTO, we expand the scope of the dataset to include pre-grant patent applications and patent re-assignment information. This addition allows users to study patent applications regardless of grant status and to observe ownership transitions beyond those related to mergers and acquisitions. A special thanks and appreciation go to Sanskriti Purohit and Ron Rabi for their diligent work and dedication to this effort. The dataset is freely available under the O-UDA-1.0 License, permitting unrestricted use for research and commercial purposes. We request that users provide proper citations when utilizing the dataset. The license also allows for the creation of derivative datasets based on DISCERN, with the condition that creators ask their downstream users to cite the original authors appropriately. If you use the data, please add these citations: 1. Arora, A., Belenzon, S., Cioaca, L., Sheer, L, Shin, H.M. & Shvadron, D. (2024). DISCERN 2.0: Duke Innovation & SCientific Enterprises Research Network [Dataset]. In Zenodo (CERN European Organization for Nuclear Research). https://doi.org/10.5281/zenodo.3594642 2. Arora, A., Belenzon, S., Cioaca, L., Sheer, L, & Shvadron, D. (2024). Back to the Future: Are Big Firms Regaining their Scientific and Technological Dominance? Evidence from DISCERN 2.0 (available soon)

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.004
metaresearch head score (Gemma)0.002
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Science and technology studies, Scholarly communication, Insufficient payload (model declined to judge)
Consensus categoriesInsufficient payload (model declined to judge)
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: Not applicable
GenreCandidate signal: Dataset · Consensus signal: Dataset
Teacher disagreement score0.187
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0040.002
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0020.007
Science and technology studies0.0060.001
Scholarly communication0.0150.001
Open science0.0030.008
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0340.220

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.123
GPT teacher head0.320
Teacher spread0.196 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; both teacher heads agree on what is shown here.

Study designNot applicable
Domainnot available
GenreDataset

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations2
Published2024
Admission routes1
Has abstractyes

Explore more

Same venueZenodo (CERN European Organization for Nuclear Research)Same topicBig Data and Business IntelligenceFrench-language works237,207