Text & Data Mining (TDM) Policy

This TDM Policy governs automated text and data mining of ESS Press content. It applies to the book line (books, monographs, chapters, and conference proceedings). For journals, see the rules published on the journals’ websites.

1) Principles

1.1. ESS Press supports research-oriented TDM of open content, provided that attribution, third-party rights, and platform stability are respected.

1.2. The purpose is to enable lawful and secure mining access with minimal administrative barriers.

2) What is permitted without a separate agreement

2.1. TDM of Open Access content published under CC BY / CC0, provided that attribution and the DOI for each used unit are included.

2.2. Extraction of “non-consumptive” facts (n-grams, frequencies, dictionaries, vector representations), where the resulting data do not allow reconstruction of substantial parts of the original text/images.

2.3. Temporary caching for the duration of analysis, with deletion of working copies after completion.

3) When coordination is required

3.1. Large-scale automated harvesting (crawling an entire catalogue/series; >N requests per minute) requires coordination of technical parameters, fair-use rate limits, and time windows.

3.2. Access to non-OA materials (volumes restricted by contract) is permitted only under separate agreements/licenses.

3.3. Use of full texts in subsequent dataset distribution (republication of VoR/AAM) requires a separate license. In most cases, secondary datasets containing factual or vector representations are permitted, rather than distributing the source texts.

4) Technical conditions and fair use

4.1. Comply with robots.txt and Cache-Control headers. Do not bypass technical protection measures.

4.2. Use a clear user-agent identification and provide a contact (email/project URL) in headers/requests.

4.3. Recommended default limits (unless otherwise agreed): ≤ 1 request/second and ≤ 3,000 records/day; parallelism should be limited.

4.4. Prefer official exports/metadata sources (Crossref, catalogue CSV/JSON, OAI-PMH, where available).

5) Attribution and reproducibility

5.1. In outputs derived from TDM, cite: (a) ESS Press as the source, (b) date range/corpus definition, (c) exact tool/script versions, (d) DOI links to analyzed works (where individually referenced), and (e) the content license.

5.2. When publishing derived datasets, include a README describing cleaning/normalization methods, field definitions, and license constraints.

6) Rights and restrictions

6.1. Authors’ moral rights must be respected (authorship and reputational integrity). It is prohibited to attribute to authors claims or conclusions that are not present in the original works.

6.2. Respect third-party rights embedded in materials (photos, maps, tables with restrictions).

6.3. Republishing full VoR texts (PDF/HTML) as part of a dataset is prohibited unless permitted by the applicable license/contract.

7) Confidentiality and legality

7.1. Do not perform TDM on personal data without a lawful basis and de-identification safeguards (see the Privacy Policy and Data Availability).

7.2. Comply with national and international copyright, database, and information security laws.

8) Relationship to other documents

This TDM Policy is applied together with: Author Rights and Licensing, Open Access & Self-Archiving, the Privacy Policy, and Publication Ethics.