Decision Tree

First check the text for the typical characteristics in the section Quickstart. If none applies, continue to the Decision Tree below. To see the details on genre categories, click on their labels or go to the section Categories Explained.

Quickstart

Do any of the following genre type characteristics apply?

• The text is not suitable for annotation (encoding issues, not in target language, machine translation or generated text, HTML source code, no full sentences, multiple different texts, intertwined texts, too long - longer than 5,000 words, too short/incoherent) - Not Suitable
• Form: verses - Prose/Lyrical
• Form: (movie/drama) script - Other
• Content: abstract, methods, results, discussion, conclusions and/or references; scientific terms - this is a research article which goes under Information/Explanation
• Content: preparation of food/drinks - this is a recipe, which goes under Instruction
• Content: conversation (between multiple people) in form of comments - Forum
• Language: legal terms - Legal
• Purpose: to invite the reader to an event/action - this is an invitation, which goes under Promotion
• Purpose: literary purpose - the main purpose is to give pleasure to the reader (and the text is not in a form of a script) - Prose/Lyrical

If not, continue to the Decision Tree below.

Decision Tree

1. Objectiveness

Is the main purpose to objectively inform the reader? Does the text contain no or almost no subjective adjectives or adverbs (already, yet, only), words that convey certainty (surely, certainly) etc.?

• Yes, it is a report on a recent event - News
• Yes, it instructs the readers on how to do something - Instruction
• Yes, but it is none of the above - Information/Explanation
• No/Not sure - go to the question 2. Subjective text

2. Subjective text

What is the purpose of the text?

• To subjectively report on a recent event - News
• To promote/sell products or services (a company) - Promotion
• To convey opinion or subjectively narrate personal experiences (no commercial purpose) - Opinion/Argumentation
• To convey opinion on a product based on personal experiences - this is a review, which goes under Opinion/Argumentation
• The purpose of the text is indiscernible or the text is not covered by other categories - Other. This label also includes interviews, correspondence (letter-like form (greeting, complimentary close etc.) adressed to a specific recipient), script, call for papers, etc.

Categories Explained

Table of Contents:

• Description of Labels
• Not Suitable
• Information/Explanation
• Instruction
• Legal
• Forum
• News
• Opinion/Argumentation
• Promotion
• Prose/Lyrical
• Other

Description of Labels

Not Suitable

If the text is not suitable for annotation, please label it with one of the following categories:

• Boilerplate: contains only the information in the header of footer (such as “Cookies”)
• Encoding Issues: there are issues with encoding/OCR. Note: If the text is written without diacritics (c, s, and z instead of č, ć, š, and ž), annotate it with genres, do not discard it.
• Generated Text: a text that was automatically generated based on a form.
• HTML Source Code: contains tags that are a part of a HTML code
• Machine Translation: a text is an obvious machine translation (contains words that do not exist in the target language, wrong declensions of nouns, adjectives, adverbs, and wrong conjugation of verbs, incoherent text)
• Multiple Texts: contains intertwined texts of various genres (e.g. an article which includes a whole letter) or multiple different texts, one after another. Also includes lists of summaries, e.g., lists of excerpts from multiple texts that end with “See more”.
• Non-textual: does not consist of sentences (receipts, statistics, code, program, forms for making queries, purchases, downloading or logging in, link collection, image collection, search page, tables, calendar, lists: bibliography, list of events (agenda), directory of persons in organization or organizations, menu, dictionary, glossary, CV/resume etc.). Note: If a text consists of some lists and other non-textual elements, but includes sentences as well, do not discard it as “non-textual”.
• Not in target languge: contains a full sentence in a foreign language that is not our target language. If there are only a few foreign words in an otherwise Slovene/English/Croatian text, do not discard it.
• Too Long: longer than 5,000 words
• Too Short/Incoherent: it is not possible to ascertain genre based on the text as it is too short / the text is incoherent, makes no sense.

Go to examples of Not Suitable

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Information/Explanation

An objective text that describes or presents an event, a person, a thing, a concept etc. Its main purpose is to inform the reader about something.

This category also includes research articles: objective, clearly-structured texts which present research, use formal language and scientific terms.

Common features:

• objective/factual: contains no or almost no adjectives/adverbs that convey opinion or promote something (great, awful, high-quality)
• explanation/definition of a concept (x is …)
• enumeration

❕ If the text describes a product or a service but is not subjective nor does it address the reader - annotate it as Information/Explanation (it is an objective description), not as Promotion. See examples on the link below.

Examples: informational blog, product specification, course materials, syllabus, textbook, general information, job description, manual, horoscope, travel guide, (informational) PowerPoint slides, encyclopedia article, glossaries (textual - containing full sentences), historical article, biographical story/history, research article, abstract.

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Instruction

An objective text which instructs the readers on how to do something.

This category also includes recipes: objective texts which instruct the readers on how to prepare food or drinks.

Common features:

• multiple steps/actions
• chronological order
• 1st person plural or 2nd person
• modality (must, have to, need to, can, etc.)
• adverbial clauses of manner (in a way that), of condition (if), of time (after …)
• common words/expressions (“it is good/smart/best to …”)

❕ If a text has features of an instruction, but it is subjective (contains subjective adjectives or adverbs, words that convey certainty), annotate it as Opinion/Argumentation as the primary category, and use the secondary category Instruction.

❕ if the text informs the reader what has to be done (rules) with many modal verbs, but does not address the reader (using 1st or 2nd person) - is more administrative, then annotate it as Instruction (primary category) + Information/Explanation (secondary category). See an example on the link below.

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Legal

An objective formal text that contains legal terms and is clearly structured. The name of the text type is often included in the headline (contract, rules, amendment, general terms and conditions, etc.).

Common features:

• objective/factual: contains no or almost no adjectives/adverbs that convey opinion or promote something (great, awful, high-quality)
• legal terms (Slovene: splošni pogoji, pravica, akt, pravilnik, zakon, uredba, Uradni list, člen, pogodba, obveznost, globa, vloga)
• common words/expressions (Slovene: pridržati, biti dolžan, pisno dovoljenje, prepovedan, kaznovati, jamčiti, odgovarjati, zavezati se)
• modality (must, can, cannot, should not, need to)
• 3rd person

❕ If the text is a explanation of a law, a layman’s interpretation, which means that the topic of the text is law, but the text does not have features of Legal (it does not include legalese, very clear structure), annotate it as Information/Explanation (primary label) + Legal (secondary label). See examples on the link below.

Examples: small print, software license, rule, proclamation, terms and conditions, contracts, law, copyright notices, university regulation …

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Forum

A text in which people discuss a certain topic in form of comments.

Common features:

• multiple authors
• informal language
• subjective (the writers express their opinions)
• written in 1st person

Examples: comments, QA forum, reader/viewer responses.

Go to examples.

❕ If the text consists of only one comment, which has characteristics of a Forum, still annotate as a Forum, even though there are not multiple people who discuss the topic.

❕ If all of the comments are reviews - the authors do not discuss the topic, each just wrote a review on a product/restaurant/etc. in a form of a comment, annotate the text as Forum (primary category) and Opinion/Argumentation (secondary category). See examples of such text on the link above.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

News

A text which reports on an event recent at the time of writing or coming in the near future.

Also includes announcements - objective texts which notify the readers about new circumstances, asking them to act accordingly. Unlike invitation (promotion) there is no intent to sell/promote – it is objective and factual.

Common features:

• topic: recent event/activities
• adverbs/adverbial clauses of time and/or place (dates, places)
• many proper nouns
• direct or reported speech
• past tense
• common words/expressions, typical for announcements (Slovene: “obvestilo, pozivati, obveščati, prositi”, Croatian: “obavijest, obaviještavamo, molimo vas”)

❕ If the text has the form of an article that could be published in the news/in a magazine, and includes a lot of reported speech (he said, “…”), it can be annotated as News, even if it does not report on a concrete event (see an example among the examples on the link below).

❕ If the authors describe their personal experience and there are doubts whether the text belongs to News or Opinion/Argumentation, consider whether there are any dates and locations mentioned. If yes and if it seems likely that such text would be published in a newspaper or as a news on a website of an organisation, annotate it as News. See more details and examples here.

❕ If the text reports on an event/product, but is very promotional, annotate it as Promotion + News (use Promotion as the primary label and News as the secondary label).

Examples: news report, sports report, travel blog, reportage, police report, notification (error message), announcement (“we would like to inform you”).

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Opinion/Argumentation

A subjective text in which the authors convey their opinion or narrate their experience. It includes promotion of an ideology and other non-commercial causes. In contrast to Promotion, the main purpose of the text is not promoting a product, a service or an event. This genre includes a subjective narration of a personal experience as well.

This category also covers reviews: texts in which authors evaluate a certain entity (a publication, a product, or a service, such as a movie, a video game, a musical composition, or a book) by endorsing or criticizing it. They usually report their personal experience with the product. It can be differentiated from the promotional text by the usage of 1st person and conveying personal experience with the product.

Common features:

• subjective: contains adjectives/adverbs that convey opinion (happy, great, aweful)
• words that convey (un)certainty (certainly, surely; Slovene: (za)gotovo, seveda, res je, dejansko, jasno)
• Slovene: Slovene particle (členek – že, še, le, tudi, sploh …)
• 1st person
• exclamation marks

❕ If the authors describe their personal experience and there are doubts whether the text belongs to News or Opinion/Argumentation, consider whether there are any dates and locations mentioned. If not and if it seems unlikely that such text would be published on a news site (it looks more like a blog), annotate it as Opinion/Argumentation. See more details and examples here.

❕ If the text is written in 1st person, that does not necessarily mean that it is Opinion/Argumentation - if the text is objective, it could be Information/Explanation (see an example on a link below).

❕ If the reviews are written in a form of comments (e.g., review comments in e-shops), annotate the text as Forum (primary category) and Opinion/Argumentation (secondary category). See examples of such text here.

❕ Use this category for texts in which the authors convey their personal experience with the product. If the text compares multiple products, but it is not based on personal experience, annotate it as Promotion (primary label) + Opinion/Argumentation (secondary category). Go to examples.

❕ If the text consists of quotes of people providing their feedback (reporting personal experience with a product/service), we annotate it as Opinion/Argumentation (see an example on a link below).

❔ How to differentiate between Opinion/Argumentation and Promotion? See more details here.

Examples: blog (personal blog, travel blog), editorial, advice, letter to editor, persuasive article or essay, formal speech, pamphlet, political propaganda, columns, political manifesto, reviews.

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Promotion

A subjective text intended to sell or promote an event, an action (e.g., collecting donations), services of a company, a product, a movie, an application, a tool, or an accommodation, housing, etc. It addresses the readers, often trying to convince them to participate in something or buy something.

Common features:

• subjective: contains adjectives/adverbs that promote something (high-quality, perfect, amazing, extraordinary, comfortable, extremely, very)
• comparative and superlative forms of adjectives and adverbs (the best, the greatest, the cheapest)
• addressing the reader (usage of 2nd person)
• exclamation marks
• common words/expressions (Slovene: ugodnost, popust, spletna trgovina, koristen, čudovit, priročen, posebnost, nič več, rešitev, na srečo, izboljšati, kakovost, želje, dolga življenjska doba; strokovnjaki, specialisti, partner, stranke, naročniki, kupci, izkušnje, storitve, nabor storitev, servis, potrebe, želje, rešitve, koristi, fleksibilnost, inovativnost, učinkovitost, strokovnost, pomagati, izvajati, svetovati, ponujati, zagotavljati, obvladati, po vaši meri, celosten, konkurenčen, hitro, natančno, profesionalno)

❕ Even if the text promotes a free product, it is still regarded as promotion (it promotes for instance the benefits of eating beans) - if the author writes positively about something that could be bought and the text could appear under a product in a e-shop, this is more of a Promotion than Information/Explanation - annotate it as Promotion.

❕ If the text describes a product or a service but is not subjective nor does it address the reader - annotate it as Information/Explanation (it is an objective description). See such instances here.

❕ If the text is an review (the author conveys her/his personal experience with the product), it is regarded as Opinion/Argumentation and not Promotion, although it could have some promotional features.

❕ If the text compares multiple products, but it is not based on personal experience, annotate it as Promotion (primary label) + Opinion/Argumentation (secondary category). Go to examples.

❕ Annotate the text as Promotion if it explicitly invites the readers (addresses the reader) to an event or is very subjective. Otherwise, if the text solely reports that an event is going to happen, does not address the readers and it does not seem that the author is the organiser of the event, annotate it as News. See an instance of such text here.

❔ How to differentiate between Opinion/Argumentation and Promotion? See more details here.

Examples: advertisement, description with intent to sell, e-shop, invitation, a product presentation page, an e-shop catalogue page, a company landing page.

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Promotion + News

If a text reports on an event/product, which means that it has characteristics of the News genre, but is very promotional (so promotional that parts of the article could be used in an advertisement, e.g. “Cannon - Revolutionary. The best in the world. A perfect novelty.”), annotate it as Promotion (primary label) and News (secondary label).

Go to examples.

❔ How to differentiate between News and a Promotion + News text? If the text is a report on an event, the default choice should be News, unless it includes promotional slogans which could appear in an advertisement as well, then use Promotion + News labels. See an example among the examples on the link above.

Prose/Lyrical

A literary text that is deemed to have no other practical purpose than to give pleasure to the reader. Often the author pays attention to the aesthetic appearance of the text. It can be considered as art. This category includes:

• lyrical texts: texts that are in a form of verses,
• prose texts: literary running texts that consist of paragraphs.

❕ This category includes jokes.

Examples: lyrics, poems, prayers, short stories, novels.

Go to examples.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.

Other

A text that is appropriate for annotation but has no clear purpose or tangible features based on which it could be annotated - a non-literary text, which does not fall under any of the categories based on its purpose.

Texts that belong to the category Other:

• quiz
• survey
• introduction into the magazine’s content, introduction to a video
• summary of a movie/book
• speech (if it cannot fit under Opinion/Argumentation)
• sermon
• user’s consent for data sharing
• texts inviting the readers to participate in a giveaway, to submit their proposals or fill out a form
• exam, exercise, worksheet
• interview
• correspondence (letter-like form (greeting, complimentary close etc.) adressed to a specific recipient)
• script
• call for papers

Go to examples of these texts.

If possible, when annotating the text as Other, add a note to which new genre it could belong. If there are many instances of this genre, it can become a new genre category.

Go back to the Decision Tree or to the Table of Contents of Categories Explained.