Pioneering the way we read and learn. That’s what L-Pub is all about.
Our core mission at L-Pub is to provide content owners with standardized tools to add meaningful value to their digital products. If you are a content owner, your business is based on delivering quality. L-Pub can help you significantly deepen the quality of your digital content by:
- enabling in-context retrieval of information relevant to your readers
- allowing you to curate that information
- implementing sophisticated, user-driven functionality
Our powerful text-analysis and annotation tools combined with a core competence in termbase structuring enable a highly personalized reading experience with curated, in-context information – and, if needed, automatically generated instructional materials. Our standardized processes ensure a low cost for content owners with a high value add for your customers. Watch our video for a quick overview of what we can do for you:
L-Pub developed its core technologies to serve the field of language learning. We believe long-term language learning can be significantly improved by reading stories, supported by learner-focused technologies and tips embedded within the texts.
We started with German as a foreign language, since Germany is the country we are based in. Together with our partners Ernst Klett Sprachen and the TU of Darmstadt, we are currently further developing this core offer to cover other world languages, as well. Our dictionary content is licensed from PONS.
L-Pub was founded in 2015 as a GmbH (roughly equivalent to a Ltd.) by David P. Steel, Leonore Kleinkauf, Battista Vailati, Franck Valentin, Marco Bellavia, Urs M. Fischer and steelecht GmbH.
L-Pub operates the following websites:
This website. Our B2B offer to publishers and other content owners.
Our brand for language learning L-Books. This website covers the overarching concept, linking to language-specific offers on other pages.
The website for people learning and teaching German, where they can purchase and read L-Books as well as access further German learning resources.
Free mini-stories for language learners. We broadcast one chapter every day. They can purchase our learning tools if they want to practice the words in the stories in the same context.
Our brand for learning German vocabulary in the context of sample sentences. vobot is our first market-ready app, on top of which our other technology is being built.
The core concept behind L-Books is to provide in-context information without having to leave the page. We achieve this by using natural language processing in connection with a curated termbase. This website demonstrates this concept by offering in-context pop-ups for specific underlined words (this website’s own specific terminology).
Here is a full list of all the terminology used on this website:
An invisible digital stamp on a word or phrase in a text that allows us to relate it to other information. For instance, if we annotate a word as a verb in the past tense, we can link it with a conjugation table and sample sentences with the same meaning of the word to assist in language learning.
automatically generated instructional materials
Once L-Pub has digitally analyzed and annotated a text, any part thereof can be turned into a series of exercises for learning words or concepts. In the text, the learner or teacher simply has to select the word or concept to be learned and, by excluding that word from the sentence, the materials will automatically be generated. One example of such a series of exercises begins with multiple choice (recognition) and culminates with full spelling of the word without hints (recollection). Learners receive immediate feedback to all answers and can follow their progress as they work through the increasingly difficult exercises.
Many words have different meanings, such as “break” which may be a verb meaning “destroy” or a noun meaning “pause”. When we can tell readers exactly which meaning of a word is correct, we have disambiguated it for them.
A dictionary has a entry for each unique word or phrase it contains (did and does would have the same entry: do). The same holds true for a termbase, the L-Pub version of a dictionary. The entry may include a variety of types of information, from definition and translation to learning tips and diagrams. The entry is everything associated with the word in the termbase.
An L-Book adds an additional layer of information and interactivity to a book, which content owners can charge their customers extra for. We recommend charging 25%-50% more than the usually listed price of the same title as an e-book.
For the proprietary software that lets customers read and interact with an L-Book, a fixed percentage of net turnover from sales is due to L-Pub. Net turnover from sales is the amount left after sales tax, shop fees and distributor commissions have been deducted.
We define a page as 300 words. Since every e-reader renders pages differently depending on the user’s settings, we have chosen a page length which roughly corresponds with that of a typical paper book.
Every title that will converted into an L-Book needs to be entered into the L-Pub system, including definition of usage rights, payment terms and numerous technical parameters. A set fee is charged for this independent of the book’s size and complexity.
Software as a Service is a type of software offer that is typically cloud-based, i.e. not installed on your own computer but accessible via a web application or at least updated regularly via an internet connection. Users typically pay for the software on a subscription basis and always have access to the latest version, without having to worry about updates. In many cases, SaaS may also be refered to as a “managed service”.
This is when L-Pub uses its proprietary software to gain insights into a text. As a result of the analysis, a number of parameters will be measured and included in the text analysis report, such as number of words in the text, number of unique words, percentage of difficult and unknown words, average sentence length as well as other details about a text’s complexity.
Translation is generally considered the direct replication of a text from one langauge into another. Often exact translations are not possible or do not sound natural in the next language.
This is a software that a user accesses via a browser. So instead of starting an application on their computer, the user types in a web address and accesses the service there. A web application is usually offered as a subscription-based service and requires the user to login.
An “Application Programming Interface” is a way for different chunks of code to speak with each other. This is especially useful if applications by two different providers are supposed to interact with each other, e.g. a translation software with a dictionary software.
CEFR learning level
The “Common European Framework of Reference for Languages” defines different levels of learning proficiency based on a scale beginning with A1 for beginners up to C2 for nearly fluent speakers. The CEFR system is often used to help teachers and learners identify books appropriate for their learning level. Various officially recognized tests are also offered for learners to demonstrate their language level according to CEFR, e.g. to future employers.
Instead of defining the path a learner should take, discovery or “personalized” learning lets them follow their own natural curiosity and thrist for knowledge. This way, learners maintain a high level of motivation and keep learning, often without even noticing. Discovery learning is best if accompanied by traditional instruction with a teacher and other learners, in order to ensure important concepts are not missed. Teachers may also prepare learners for discovering learning phases by cueing key learning topics in advance.
A small popup window that appears when a reader selects a word in a text. It provides additional information without covering up the original context.
A collection of terms and/or reference content which a learner wants to commit to memory. The list is typically personalized, as only the terms are added which are relevant to the learner. The terms can be practiced at any time in a series of automatically generated exercises found in the learning mode of an L-Book.
The phenomenon of repeatedly practicing memorization of a word or concept within a short period of time. While the knowledge may remain accessible to the learner’s mid-term memory (e.g. for a text the next morning), it usually fades away quickly and will not be committed to long-term memory. Also known as massed repetition.
Whenever a reader encounters a word or phrase they would like to commit to memory, they simply have to press the plus symbol (+) in the popup to add it to their personalized learning list.
The concept of “return on investment” (ROI) is to identify the point at which the amount I have paid for a product or sevice will be covered by the additional income I can generate thanks to that product or service. For instance, I pay 100 euros for an advertisement and as a direct result I can sell my product and make a profit of 100 euros within one week. That means the ROI for the advertisement is reached after in one week. All profit after that is real profit.
A dictionary that is set up as a database with a list of terms and corresponding information attached to that term, such as a definition, translation, synonym, cross reference or diagram.
This is when digital texts offer additional features, besides just the possibility to read on a screen. Text enrichment allows readers to interact with the text, e.g. to access reference content, which may include definitions, footnotes, sounds or pictures.
Technische Universität (type of German university with a predominantly scientific, engineering or technological focus)
This is when a software is reproduced with nearly the same functionalities but is given a different “label”. For instance, L-Pub has a branded software that a company would like to sell as their own. The company pays L-Pub a license and can then sell our software as their own.
Any text used in foreign language acquisition which was not written for that purpose, i.e. not with a learning audience in mind. Almost any type of text may be considered authentic literature: novels, websites, brochures, lyrics, etc.
The price of converting a text into an L-Book. It is calculated based on a number of variables, starting with the length of the book (number of words) and taking into consideration factors like volume discounts and degree of customization requested.
When confronted with a multiple choice question, in many cases it is easy for a learner to deduce the correct answer simply by excluding obvious wrong answers. If the possible answers are similar, the incorrect options are refered to as distractors – they are placed on purpose to make sure the learner is really considering the possible answers and not just using deduction.
Even if a book is printed on paper or bought outside of an L-Reader ecosystem, if it is “L-Book enabled”, readers can still benefit from the additional functionality. Readers find out about the different ways to access the L-Book features via the website L-Book.info
An L-Book Pro has two main modes, the reading and the learning mode. A reader can decide at any time to interrupt their reading and switch to the learning mode and practice the terms they have collected while reading. An L-Book Lite only has the learning mode (without the reading mode).
natural language processing
This is a field of computer science which is dedicated to the machine-based analysis of language. The insights gained from texts analyzed using natural language processing (NLP) are wide and varied, but include the ability to recognize word inflections, grammar phenomena and sentence patterns. L-Pub uses these insights to bring automation into the annotation and enrichment of digital texts.
Before an L-Book can be created, a pre-analysis is necessary in order to determine important details about the text, inlcuding its length, number of unique words and general complexity. Based on these and other text features, L-Pub will be able to draft a precise quote. The pre-analysis report can also be very valuable on its own for content owners who want to gain deeper insights into the nature of the language in their texts.
Text and/or image provided to readers when they click or press on a word in a text. The reference content provides meaningful additional information, such as a definition, directly in the context in which the reader needs it.
Many sectors and companies have specific words they use to describe certain objects, concepts or processes. A terminology serves to assure that everyone using these words understands their meaning and applies them correctly. A terminology is often also multi-lingual, especially for companies selling their products internationally.
The process by which language is made readable for a computer. To make language understandable and procesable, the computer first has to break it up into pieces of data (e.g. sentences and words) each of which is referred to as a token. Tokenization is an important prestep to text analysis.
Also known as a “lemma”, each unique word is the base form of a word. No matter how many times the word appears in a text or how many different ways it is spelled (plural forms or other inflections), it counts as one unique word for that text.