The PMHR 2011 workshop is held in conjunction with the ACM Hypertext 2011 (HT2011) Conference.
The aims of this workshop are twofold:
- to set the scene in this challenging area, allowing the different communities engaged in related research topics to meet and to determine a program of actions to undertake
- to set up a strategy to follow for the evaluation of PMHR systems, so to define the collection of hypertext documents to use to evaluate such type of systems together with the metrics to use.
The workshop results will be of use in the design of personalised tools that can help end-users fully benefit from the use of distributed multilingual hypertext content.
Search engines have traditionally followed a “one size fits all” paradigm and returned the same results for all users. They do not adapt to the user, the domain, or the search context. Thus, the search process and the number and type of results returned are not tailored to the individual user or her/his search situation. Personalised hypertext retrieval is concerned with adapting the search process to the user’s needs. This includes adapting the system, the query-document similarity metrics, the search results, and their presentation to an individual user. The personalisation process can be based on models of the user, the domain, and the search context, but no standard representation or resources have evolved to-date.
It can often be the case that non-native English speakers suffer limited or restricted online experiences as typically the majority of web content is still authored in English. Machine translated versions of content may be generated for some languages, but this is not always the case. This results in greatly restricted content collections.
This workshop explores the use of multilingual hypertext retrieval technologies and adaptive personalisation techniques to enable end-users to conduct searches in their native language, but receive results collated from content collections in a variety of languages, all tailored for consumption by the end-user.
The PMHR workshop aims to promote the exchange of ideas between researchers working on hypertext and adaptive hypertexts, cross-lingual information retrieval, personalised search, personalisation for Web and hypertexts, and recommender systems. The workshop will have a specific focus on research in user modelling and multilingual personalisation for hypertext retrieval. In addition, submissions which focus on non-English data or research with a clear application in a multilingual scenario are welcome.
Topics include, but are not limited to:
- Multilingual semantic search and intelligent information retrieval, extraction and filtering (e.g. How does a multilingual setting affect personalised hypertext IR?)
- Multilingualism in semantic search, semantic web, or in context-aware and semantic recommender systems
- Recommender systems, adaptation engines, and algorithms for personalised multilingual IR
- User modelling and adaptation (e.g. creation and exploitation of individual or stereotypical user profiles)
- Content personalisation and personalised result presentation (e.g. result presentation beyond the ranked list to enable users to fully benefit from the semantics carried by the hypertext structure)
- Domain modelling (e.g. adaptation to different domains)
- Creating relevant linguistic resources (processing user models, query logs or forum postings); privacy issues
- External knowledge resources for personalised multilingual hypertext IR (e.g. ontologies)
- Personalisation of multilingual tools
- Tools and methods for bilingual search
- Evaluation methodologies and metrics for personalised hypertext IR (e.g. How can the evaluation gap between IR/multilingual text retrieval and research disciplines such as adaptive hypermedia be bridged?)