ARCOMEM aims to transform archives into collective memories more integrated with their communities by exploiting social web and crowdsourcing. Advanced research is required, including analyzing social web content and community structures; detecting and consolidating events and related entities within and between archives; detecting perspectives, opinions, and sentiments expressed on the web; detecting duplicate content while maintaining diversity; intelligent decision support combining social web analysis and other inputs; advanced crawling integrating event-centered and entity-centered strategies; and approaches for long-term interpretability and semantic preservation of archive content and original contexts.

1. arcomem_lay01.indd 1 05.07.11 16:48

2. ARCOMEM’s aim is to help transform
Social Web analysis and Web mining, which includes
archives into collective memories that effective methods for the analysis of Social Web content,
are more tightly integrated with their analysis of community structures, discovery of evidence
for content appraisal, analysis of trust and provenance,
community of users and exploit Social and scalability of analysis methods.
Web and the wisdom of crowds to
Event detection and consolidation, which includes informa-
make Web archiving a more selective tion extraction technologies for detection of events and
and meaning-based process. for detecting entities related to events; methods for con-
solidating event, entity and topic information within and
between archives, models for events, covering different
In order to reach its ambitious scien- levels of granularity, and their relations.
tific and technological objectives, Perspective, Opinion and Sentiment detection, which
advanced research is required in the includes scalable methods for detecting and analysing
opinions, perspectives taken, and sentiments expressed
ARCOMEM project. This includes in the Web and especially Social Web content.
research in the following areas >>>>>
Concise content purging, which includes detection of
duplicates and near-duplicates and an adequate re-
flection of content diversity with respect to textual
content, images, and opinions.
Intelligent adaptive decision support, which includes
methods for combining and reasoning about input from
social Web analysis, diversity and coverage aspects,
extracted information, domain knowledge and
heuristics, etc.; methods for adapting the
decision strategies to inputs received.
Advanced Web Crawling, which
includes the integration of event-
centred and entity-centred strate-
gies, the use of social Web clues in
crawling decisions and methods
for translating by example and
descriptive crawling specifications
into crawling strategies.
Approaches for “semantic preser-
vation”, which includes methods for
enabling longterm interpretability
of the archive content; methods
for preserving the original context
of perception and discourse in a
semantic way; methods for deal-
ing with evolution on the semantic
layer.
arcomem_lay01.indd 2 05.07.11 16:48