Tuukka Ruotsalo

Directing exploratory search: reinforcement learning from user interactions with keywords

Techniques for both exploratory and known item search tend to direct only to more specific subtopics or individual documents, as opposed to allowing directing the exploration of the information space. We present an interactive information retrieval system that combines Reinforcement Learning techniques along with a novel user interface design to allow active engagement of users in directing the search. Users can directly manipulate document features (keywords) to indicate their interests and Reinforcement Learning is used to model the user by allowing the system to trade off between exploration and exploitation. This gives users the opportunity to more effectively direct their search nearer, further and following a direction. A task-based user study conducted with 20 participants comparing our system to a traditional query-based baseline indicates that our system significantly improves the effectiveness of information retrieval by providing access to more relevant and novel information without having to spend more time acquiring the information.

Interactive intent modeling: information discovery beyond search

The system should let users incrementally direct their search toward relevant, though not initially obvious, information.

How to deal with massively heterogeneous cultural heritage data: lessons learned in CultureSampo

This paper presents the CultureSampo system from the viewpoint of publishing heterogeneous linked data as a service. Discussed are the problems of converting legacy data into linked data, as well as the challenge of making the massively heterogeneous yet interlinked cultural heritage content interoperable on a semantic level. In the approach described, the data is published not only for human use, but also as intelligent services for other computer systems that can then provide interfaces of their own for the linked data. As a concrete use case of using CultureSampo as a service, the BookSampo system for publishing Finnish fiction literature on the semantic web is presented.

A method for determining ontology-based semantic relevance

The semantic web is based on ontologies and metadata that indexes resources using ontologies. This indexing is called annotation. Ontology based information retrieval is an operation that matches the relevance of an annotation or a user generated query against an ontology-based knowledge-base. Typically systems utilising ontology-based knowledge-bases are semantic portals that provide search facilities over the annotations. Handling large answer sets require effective methods to rank the search results based on relevance to the query or annotation. A method for determining such relevance is a pre-requisite for effective ontology-based information retrieval. This paper presents a method for determining relevance between two annotations. The method considers essential features of domain ontologies and RDF(S) languages to support determining this relevance. As a novel use case, the method was used to implement a knowledgebased recommendation system. A user study showing promising results was conducted.

Predicting term-relevance from brain signals

Term-Relevance Prediction from Brain Signals (TRPB) is proposed to automatically detect relevance of text information directly from brain signals. An experiment with forty participants was conducted to record neural activity of participants while providing relevance judgments to text stimuli for a given topic. High-precision scientific equipment was used to quantify neural activity across 32 electroencephalography (EEG) channels. A classifier based on a multi-view EEG feature representation showed improvement up to 17% in relevance prediction based on brain signals alone. Relevance was also associated with brain activity with significant changes in certain brain areas. Consequently, TRPB is based on changes identified in specific brain areas and does not require user-specific training or calibration. Hence, relevance predictions can be conducted for unseen content and unseen participants. As an application of TRPB we demonstrate a high-precision variant of the classifier that constructs sets of relevant terms for a given unknown topic of interest. Our research shows that detecting relevance from brain signals is possible and allows the acquisition of relevance judgments without a need to observe any other user interaction. This suggests that TRPB could be used in combination or as an alternative for conventional implicit feedback signals, such as dwell time or click-through activity.

CultureSampo: A National Publication System of Cultural Heritage on the Semantic Web 2.0

CultureSampo is an application demonstration of a national level publication system of cultural heritage contents on the Web, based on ideas and technologies of the Semantic (Web and) Web 2.0. On the semantic side, the system presents new solutions to interoperability problems of dealing with multiple ontologies of different domains, and to problems of integrating multiple metadata schemas and cross-domain content into a homogeneous semantic portal. A novelty of the system is to use semantic models based on events and narrative process descriptions for modeling and visualizing cultural phenomena, and for semantic recommendations. On the Web 2.0 side, CultureSampo proposes and demonstrates a content creation process for collaborative, distributed ontology and content development including different memory organizations and citizens. The system provides the cultural heritage contents to end-users in a new way through multiple (nine) thematic perspectives, based on semantic visualizations. Furthermore, CultureSampo services are available for external web-applications to use through semantic AJAX widgets.

An event-based approach for semantic metadata interoperability

This paper presents a method for making metadata conforming to heterogeneous schemas semantically interoperable. The idea is to make the knowledge embedded in the schema structures interoperable and explicit by transforming the schemas into a shared, event-based representation of knowledge about the real world. This enables and simplifies accurate reasoning services such as cross-domain semantic search, browsing, and recommending. A case study of transforming three different schemas and datasets is presented. An implemented knowledge-based recommender system utilizing the results in the semantic portal CULTURESAMPO was found useful in a preliminary user study.

Designing for Exploratory Search on Touch Devices

Exploratory search confront users with challenges in expressing search intents as the current search interfaces require investigating result listings to identify search directions, iterative typing, and reformulating queries. We present the design of Exploration Wall, a touch-based search user interface that allows incremental exploration and sense-making of large information spaces by combining entity search, flexible use of result entities as query parameters, and spatial configuration of search streams that are visualized for interaction. Entities can be flexibly reused to modify and create new search streams, and manipulated to inspect their relationships with other entities. Data comprising of task-based experiments comparing Exploration Wall with conventional search user interface indicate that Exploration Wall achieves significantly improved recall for exploratory search tasks while preserving precision. Subjective feedback supports our design choices and indicates improved user satisfaction and engagement. Our findings can help to design user interfaces that can effectively support exploratory search on touch devices.

Elements of a National SemanticWeb Infrastructure--Case Study Finland on the Semantic Web

This article presents the vision and results of creating the basis for a national semantic Web content infrastructure in Finland in 2003-2007. The main elements of the infrastructure are shared and open metadata schemas, core ontologies, and public ontology services. Several practical applications testing and demonstrating the usefulness of the infrastructure are overviewed in the fields of eculture, ehealth, egovernment, elearning, and ecommerce.

Knowledge-Based Linguistic Annotation of Digital Cultural Heritage Collections

A method for automatically annotating objects in digital cultural heritage collections uses structured vocabulary concepts and their metadata schema roles.