Syllabs was born from a technological vision to meet information handling requirements:

• 1- Semantic analysis can and must play a major role in our approach to information
• 2- The generic semantic analyses that are generally offered by current players in the field cannot operate correctly when dealing with specific needs
• 3- The quantity of information to be processed is often very high and therefore high speed parsers are necessary
• 4- Information evolves very rapidly and the parsers that have been developed must therefore be easily and incrementally modifiable
• 5- Semantic analysis is only a part of a much greater whole and it is thus necessary to place it into larger solutions.

From the very start, Syllabs’ purpose has been to develop semantic technology that is easy to specialize, fast, robust and integrated into a large value chain.

To bring its vision to fruition, Syllabs developed over three phases:

2006-2009: Syllabs started off as a Private Research Laboratory, providing consultancy services and expertise to numerous companies and public laboratories, while at the same time developing its own technology. From this period, Syllabs developed strong dedication to satisfying its clients and the ability to perceive, understand and adapt to its clients’ specific needs. Syllabs gained much experience during this period in all domains related to semantics and earned the status of Technological Research Organization.

2009-2011: As a technological startup offering very innovative, high performance products, Syllabs concentrated its efforts on its technology, selling semantic analyses mainly through its SaaS platform which enables the processing of very high quantities of information in a very short time. In this way, Syllabs developed technologies for entity extraction, specialized opinion analysis, terminological extraction, etc. and process millions of requests each day.

Since 2011: Syllabs’ product portfolio enables it to offer the world unique vertical solutions by combining its three major technologies i.e. web mining, text mining and generating unique textual content of human quality.

Innovation is deeply rooted in Syllabs’ DNA and it would like to thank the following organizations dedicated to research and development for their support:

• > The economic development agency of Paris: Paris Développement (Invest in Paris)
• > The European Commission, through the PCRD programmes
• > The ANR (French National Research Agency)
• > The incubator Agoranov
• > OSÉO
• > The ANRT (French National Technology Research Association)
• > The Ministry of Research (status of Innovative Young Enterprise).

Syllabs would also like to thank its partners who have been there throughout its research projects and who have enabled it to develop its technologies:

• > Public Laboratories: ALPAGE, AFP, CIM-Paris3/ENS, CNAM, CTS (Leeds), CRIS, ELICO, ESIEE, Eurecom, GERiiCO, IMS (Stuttgart), Ina, IRIT, LIA,  LiCoRN, LIF, LIG, LIMSI, LINA, LIP6, LIPN and VALORIA.
• > Enterprises: A2IA, Acapela, Advestigo, AFP, Aldebaran Robotics, Banctec, Lingua et Machina, Linkfluence, Qwam Content Intelligence, Safig/Jouve, Sinequa SAS, Sogitec Industries, Tilde (Riga), TNS Media Intelligence, Vocapia Research and Wikio Group.

• > “Cet article n’a pas été écrit par un robot” (“This article was not written by a robot”)- Stratégies n°1662 19/01/2012

If you want to work with us, please contact us or visit our job opportunities.

Syllabs Web Mining is a development environment enabling information from the web to be  searched for, collected, merged and indexed. This technology allows highly specialized vertical search engines to be created for example, that would enable users to access all available information on any one given product, place or person, etc. in just a click. Thanks to unique semantic technology, this environment allows Syllabs to adapt its crawlers to all website configurations, limiting the setting of the system parameters to just a few simple and rapid manipulations. Data collected in this way from no matter how many websites is merged and indexed within specialized search engine. This search engine can then be opened outwards to allow access to all available information.

Syllabs Text Mining offers a unique software environment based on an object oriented programing language dedicated to semantic analysis, providing a solution for the specificities of each sphere of application. This language allows Syllabs’ engineers to program specialized semantic parsers in very little time, in a highly-developed programing environment. It was also designed to process large volumes of text in very little time.Syllabs Text Mining is currently available in 10 languages: German, English, Chinese, Spanish, French, Italian, Dutch, Polish, Portuguese and Russian. Other languages will be developed soon. Certain analyses are already available on Syllabs’ SaaS platform and all possibilities remain open for other types of processing: specialized opinion analysis, extraction of contact details, following company mergers, etc.

Syllabs Content Generator enables human quality textual content to be generated automatically (e.g. generating a product description from its technical datasheet). This technology also allows existing content to be enriched (from deductions made from available information) with the aim to improve the overall user experience. Take for example the descriptions of shoes on an eCommerce site. If the product is a sandal, further information like “ideal for the summer” could be added. Or if the product was a running shoe, “for your comfort” could be added instead, customizing the text to make the product more attractive.

Added value of Syllabs’ content:

• > Syllabs’ content can be produced in several languages, without the necessity of translation (content generated directly in the language).
• > Much work is done on the variability of texts thanks to a description of the multiple possible ways to provide the same information. Due to this variability, the unicity of the texts produced can be ensured.
• > Syllabs can generate very large volumes, practically in real time (managing several hundred thousand texts per day).

ROI:

• > Better user experience: Syllabs can provide information that a technical datasheet does not provide to non-technical users. So, as not everyone knows that such and such a brand of photo lens has a great reputation, for example, adding this information to the text, helps the Internet user to make their choice.
• > Increasing traffic through search engine optimization: the presence of human quality unique texts has a direct impact on indexing.
• > Automatic update of catalogue descriptions: the arrival of new products implies large updating of content, which is not always possible manually, especially for bilingual or multilingual sites.

Examples of use: automatic generation of original texts such as product descriptions (e-Commerce) or hotels (e-Tourism).Syllabs Content Generator offers a unique text generation solution providing a solution for the specificities of each sphere of application.

The Syllabs eTourism solution brings an answer to two vital difficulties encountered by eTourism sites:

• > Getting Internet users to come to their sites
• > Informing and seducing users to entice them more easily to buy.

Syllabs knows that it is very difficult to write up or have written up by hand thousands of descriptions of hotels, restaurants or any other kind of place or site. It offers on demand catalogues of descriptions of tourist sites including:

• > Unique texts (Syllabs can generate thousands of different texts to describe a single site)
• > Hotel descriptions using information it gathers from the whole web
• > Summaries of the opinions of Internet users.

Texts provided in this way can be directly included in the pages of the website and offer a solution to both of the mentioned difficulties.

The Syllabs eCommerce solution enables commercial websites to increase both the number of visits to and the conversion rate of their sites.

Practically all eCommerce sites display the same product descriptions and bring nothing new, either to the search engine or to the Internet user. The automatic generation of descriptions offered by Syllabs enables unique texts to be created that satisfy the requirements of search engines and that bring a description to the Internet user that is not only nice to read but that could include additional information such as the brand’s reputation, or interesting facts like how a particular type of camera can have a very large display screen, etc. The texts are generated from a product datasheet and can also include a summary of the opinions of Internet users if there are any available.

No need to be a developer to use Syllabs’ solutions… but if you happen to be one…

Certain customers want to keep control of the process chain. Syllabs Web Services propose to use existing analyses to make up your own system. Syllabs can also include specialized, tailor-made web services on this platform to meet very specific semantic analysis requirements. Documentation on this can be found at the following address: http://www.api.syllabs.com

• EDyLex: Dynamic acquisition of lexical entries and their associated information. Funded by: ANR (French national research agency). Partners: Alpage (leader), AFP, LIF, LIMSI and Vocapia Research.
• Feed-ID: Qualification and enrichment of blogs for their indexing. Funded by: DGCIS. Partners: Syllabs (leader), Linkfluence and Wikio Group.
• GV-Lex: To equip a robot with the capability of expressive reading. Funded by: ANR (French national research agency). Partners: Aldebaran Robotics (leader), Acapela, LIMSI and Telecom Paris Tech.
• MéTRICC: Construction and use of comparable corpora, for translation memories, interlingual searching and multilingual categorization. Funded by: ANR (French national research agency). Partners: LINA (leader), Lingua et Machina, LiCoRN, LIG, Sinequa SAS and VALORIA.
• OTMedia: Research into the development and transformation of the media world. Funded by: ANR (French national research agency). Partners: Ina (leader), AFP, INRIA, LIA and CIM-Paris3/ENS.
• SuMACC: Learning methods to detect multimedia entities by using specific detection patterns. Funded by: ANR (French national research agency). Partners: LIA (leader), Eurecom and Wikio Group.
• TTC: Automatic generation of bilingual lexicons from comparable corpora. Funded by: European Commission (FP7). Partners: LINA (leader), CTS, Eurinnov, IMS, Sogitec Industries and Tilde.

• Blogoscopie: Analysis of opinions in blogs. Funded by: ANR (French national research agency). Partners: LINA (leader), OverBlog and Sinequa SAS.
• Piithie : Detection of plagiarism and assessment of its media impact. Funded by: ANR (French national research agency). Partners: Sinequa SAS (leader), Advestigo SA, LIA and LINA.
• RPM2: Multimedia summary. Funded by: ANR (French national research agency). Partners: Sinequa SAS (leader), Institut Eurecom, Laboratoire Informatique d’Avignon and Wikio.
• Adonis: TGE Adonis platform.

• > Guégan, M., & Loupy, C. de. (2011). RANLP 2011 – Recent Advances in Natural Language Processing.
• > Groc, Clément de, Tannier, X., & Couto, J. (2011). GrawlTCQ: Terminology and Corpora Building by Ranking Simultaneously Terms , Queries and Documents using Graph Random Walks. In Proceedings of the 6th Workshop on Graph-based Methods for Natural Language Processing, Association for Computational Linguistics (p. 37–41).
• > Blancafort, H., Heid, U., Gornostay, T., Méchoulam, C., Daille, B., & Sharoff, S. (2011). User-centred Views on Terminology Extraction Tools: Usage Scenarios and Integration into MT and CAT Tools. Paris, France: TRALOGY Conference “Translation Careers and Technologies: Convergence Points for the Future”.
• > Groc, Clément de. (2011). Lecture Notes – Data-Intensive Text Processing with MapReduce. Natural Language Processing, XXX, 1-4.
• > Groc, Clément de, Couto, J., Blancafort, H., & Loupy, C. de. (2011). Natural Language Processing.
• > Blancafort, H., & Loupy, C. de. (2011). Clues to Compare Languages for Morphosyntactic Analysis: A Study Run on Parallel Corpora and Morphosyntactic Lexicons. Lecture Notes on Artificial Intelligence.
• > Couto, J., Groc, Clément de, Guégan, M., & Loupy, C. de. (2011). Use of faceted classification in search engines. In L. Grivel (Ed.), Technologies de la connaissance et Recherche d’Information en Contexte (Knowledge technologies and in-context information searching, in French).
• > Blancafort, H., Recourcé, G., Couto, J., Sagot, B., Stern, R., & Teyssou, D. (2010). TALN 2010.
• > Blancafort San José, H., Couto, J., & Seng, S. (2010). Morphosyntactic Lexica in the OAL Framework: Towards a Formalism to Handle Spelling Variants, Compounds and Multi-word Units. EURALEX 2010. Leeuwarden, Netherlands.
• > Blancafort, H. (2010). Learning Morphology of Romance, Germanic and Slavic languages with the tool Linguistica. LREC’10 – 7th Conference on International Language Resources and Evaluation (pp. 2257-2262).
• > Blancafort, H., Daille, B., Gornostay, T., Heid, U., Méchoulam, C., & Sharoff, S. (2010). Terminology Extraction, Translation Tools and Comparable Corpora. EURALEX 2010. Leeuwarden, Netherlands.
• > Couto, J., Blancafort, H., Seng, S., Kuchmann-beauger, N., & Loupy, C. de. (2010).OAL: A NLP Architecture to Improve the Development of Linguistic Resources for NLP. LREC’10 – 7th Conference on International Language Resources and Evaluation (pp. 975-981).
• > Loupy, C. de, Guégan, M., Ayache, C., Seng, S., & Torres-Moreno, J.-M. (2010). A French Human Reference Corpus for Multi-Document Summarization and Sentence Compression. In N. (Conference C. Calzolari, K. Choukri, B. Maegaard, J. Mariani, J. Odijk, S. Piperidis, et al. (Eds.), LREC’10 – 7th Conference on International Language Resources and Evaluation. Valletta, Malta: European Language Resources Association (ELRA).
• > Masmoudi, H., Peigné, Q., De Loupy, C., Den Besten, M., & Dalle, J.-M. (2009). Fifth International Conference on Open Source Systems – OSS 2009. Skovde, Sweden.
• > Waszak, T., Loupy, C. de, & Bellot, P. (2009). Detecting and hierarchical structuring of titles in HTML documents. CORIA 2009 – Conference on Information and Application Searching.
• > Couto, J., Blancafort, H., Seng, S., Talby, A., & Loupy, C. de. (2009). OAL: a tool for the development of linguistic resources. Procesamiento del Lenguaje Natural, 43, 369-370.
• > Blancafort, H., & Loupy, C. de. (2009). Clues to Compare Languages for Morphosyntactic Analysis: A Study Run on Parallel Corpora and Morphosyntactic Lexicons. LTC’09. PoznaÅ„, Poland.
• > Loupy, C. de, Bagur, M., & Blancafort, H. (2009). TALN’2009. Senlis, France.
• > Loupy, C. de, & Gonçalves, S. (2008). EURALEX 2008 (pp. 331-337). Barcelona, Spain.
• > Blancafort, H., & Loupy, C. de. (2008). Comparing languages from vocabulary growth to inflection paradigms: A study run on parallel corpora and multilingual lexicons (estudio sobre corpus paralelo y léxicos multilingües). SEPLN 2008 – 24th Conference of the Spanish Society for Natural Language Processing (Vol. 41, pp. 113-120).
• > Cailliau, F., & Loupy, C. de. (2007). TALN’2007. Toulouse, France.

Syllabs – www.syllabs.com

Head Office: 53 bis rue Sedaine 75011 Paris

Limited liability company with a registered capital of 18,000 Euro – Paris B Trade and Company Registry N° 491 380 754

Syllabs.com and syllabs.fr websites are published by the company Syllabs. All publications hosted on these websites are protected by copyright. They can be partially reproduced once Syllabs is cited as their author and its website address is listed as http://www.syllabs.com. For any further information, please contact Syllabs by email at the email address: contact@syllabs.com.

Hosting

This site is hosted by OVH - 140 quai du Sartel, 59 100 Roubaix – www.ovh.com

Photo Credits:

• renjith krishnan / FreeDigitalPhotos.net
• Master isolated images / FreeDigitalPhotos.net
• Danilo Rizzuti / FreeDigitalPhotos.net