Project: Managing Academic Knowledge with INtegrated, collaboratIve Tools
The general objective of the MAKIN’ IT project is to develop a solid back-end data management system leading to a consistent database of academic knowledge, driven by semantic algorithms, to efficiently manage and share academic knowledge and to achieve superior recommendations for academic content based on personal profiles and user preferences._x000D__x000D_Building up and COtaining a personal digital library of scholarly articles and researching additional relevant content and authors can be a time consuming and difficult process for every researcher, academic, and student. Currently there are no services to give users automated recommendations based on the actual research context of the user (personal library, profile, and preferences) or the content of a specific paper. Recommendations are instead given based on user generated content (mostly tags or readership statistics), keyword notifications or citation counts. Additionally, stand-alone literature management software has little online functionality, is mostly unable to function on more than one computer architecture or operating system and is not integrated with databases or academic networks._x000D_The solution proposed by MAKIN’ IT would be based on the principles of a well-known, successful Web 2.0 service, namely the world’s largest social music service “Last.fm”. The idea is to create a similar service for academics, researchers, and students – a “Last.fm for research” – which would consist of two parts: First, a cross-platform desktop application (available for Windows, Mac and Linux) which helps researchers manage, share and collaboratively tag and annotate their academic papers and which anonymously generates usage statistics. Second, a website where users can discover aggregated statistics about top papers and authors, trends and charts for each discipline, paper recommendations, and introductions to people with similar research interests (see also the simplified user scenario in the annex)._x000D_In order to COtain a consistent database and to generate relevant recommendation results, the underlying data has to be clean and inter-connected. Therefore, the general objective of the project is realised by setting following specific measurable objectives:_x000D_• Extracting metadata (author(s), title, year, publication, references, etc.) from scholarly articles to establish a database of academic knowledge. This means to convert unstructured information extracted from PDF documents into a structured format. Existing algorithms will be improved and new algorithms developed which can extract metadata from PDF documents with different formats and layouts (work package 1)._x000D_• Matching data on the server by generating reliable fuzzy fingerprints or hash values based on the PDF document or the extracted metadata. These fingerprints should at the same time allow for enough flexibility to identify the same documents even if they come along in different formats (e.g. text-PDF vs. image-PDF) or with different recognition errors (work package 2)._x000D_• Developing algorithms that are able to determine, without human guidance or pre-defined target keywords, the content similarity of academic papers, which will be used for analyses and recommendations (work package 3)._x000D_• Generating an innovative automated entity disambiguation system to place author names in the context of their papers and other prominent features, hence enabling automatic distinction of authors with the same name and merging author names with different spellingsor abbreviations, and by this allowing to publish author/profile and article pages on Mendeley Web (work package 4)._x000D_• Developing a recommendation engine to connect content and people based on the user’s research interests and personal digital library of scholarly articles. In addition, the results of the research project will be tested and implemented into Mendeley Desktop and Web (work packages 5 and 6)._x000D_The specific results of the project (listed in section 2.2.3) will be integrated into the Mendeley software and web services. The developed solution will be competing on two broad markets – on the software market (here specifically the knowledge management software market) and on the online services market. Commercialization of project results will be carried out by the CO participant (see section 2.3 for detailed market analysis)._x000D_The consortium of the project will consist of three Ps: the developing company Mendeley Ltd. (Mendeley, United Kingdom), independent research organisation Competence Centre of Electronics-, Info- and Communication Technologies (ELIKO, Estonia) and Austria’s Competence Centre for Knowledge Management (Know-Center, Austria). Mendeley’s competencies lie in metadata extraction, fuzzy fingerprinting and recommendation engines. ELIKO and Know-Center will respectively deal with questions regarding semantic similarity/relatedness and named entity disambiguation and classification (see also section 2.4.1)._x000D_
| Acronym | MAKIN' IT (Reference Number: 4811) |
| Duration | 01/04/2009 - 30/09/2010 |
| Project Topic | The general objective of the MAKIN’ IT project is to develop a consistent database of academic knowledge, driven by semantic algorithms, to efficiently manage and share academic content, and to achieve superior recommendations based on personalised user profiles and user preferences. |
|
Project Results (after finalisation) |
A back-end data managment system for academic knowledge capable of converting scientific papers in PDF to plain text, extract meta-data such as authorship information or references, disambiguate meta-data, deduplicate papers based on fuzzy fingerprinting and similarity comparisons as well as content recommendation. The data managment system is an integral part of Mendeley's product._x000D__x000D_ |
| Network | Eurostars |
| Call | Eurostars Cut-Off 2 |
Project partner
| Number | Name | Role | Country |
|---|---|---|---|
| 3 | Know-Center - Competence Center for knowledge-based Applications and Systems | Partner | Austria |
| 3 | Mendeley Ltd. | Coordinator | United Kingdom |
| 3 | OÜ ELIKO Tehnoloogia Arenduskeskus / ELIKO Competence Centre in Electronics-, Info- and Communication Technologies | Partner | Estonia |