As current methods for content-based retrieval are incapable of capturing the semantics of images, we experiment with using spectral methods to infer a semantic space from user’s relevance feedback, so that our system will gradually improve its retrieval performance through accumulated user interactions. In addition to the long-term learning process, we also model the traditional approaches to query refinement using relevance feedback as a short-term learning process. The proposed shortand long-term learning frameworks have been integrated into an image retrieval system. Experimental results on a large collection of images have shown the effectiveness and robustness of our proposed algorithms.

Recently, Artificial Intelligence techniques have proved useful in helping users to handle the large amount of information on the Internet. The idea of personalized search engines, intelligent software agents, and recommender systems has been widely accepted among users who require assistance in searching, sorting, classifying, filtering and sharing this vast quantity of information. In this paper, we present a state-of-the-art taxonomy of intelligent recommender agents on the Internet. We have analyzed 37 different systems and their references and have sorted them into a list of 8 basic dimensions. These dimensions are then used to establish a taxonomy under which the systems analyzed are classified. Finally, we conclude this paper with a cross-dimensional analysis with the aim of providing a starting point for researchers to construct their own recommender system.

The Web is an obvious source of valuable information but the process of collecting, organizing and utilizing these resources is difficult. We describe CorpusBuilder, an approach for automatically generating Web-search queries for collecting documents matching a minority concept. We use the concept of text documents belonging to a minority natural language on the Web. Individual documents are automatically labeled as relevant or non-relevant using a language filter and the feedback is used to learn what query-lengths and inclusion/exclusion term-selection methods are helpful for finding previously unseen documents in the target language. Our system learns to select good query terms using a variety of term scoring methods. We find that using odds-ratio scores calculated over the documents acquired so far was one of the most consistently accurate query-generation methods. We also parameterize the query length using a Gamma distribution and present empirical results with learning methods that vary the time horizon used when learning from the results of past queries. We find that our systems performs well whether we initialize it with a whole document, or with a handful of words elicited from a user. Experiments applying the same approach to multiple languages are also presented showing that our approach generalizes well across several languages regardless of the initial conditions. 1.

The Web is a valuable source of language specific resources but the process of collecting, organizing and utilizing these resources is difficult. We describe CorpusBuilder, an approach for automatically generating Web-search queries for collecting documents in a minority language. It differs from pseudo-relevance feedback in that retrieved documents are labeled by an automatic language classifier as relevant or irrelevant, and this feedback is used to generate new queries. We experiment with various query-generation methods and query-lengths to find inclusion/exclusion terms that are helpful for retrieving documents in the target language and find that using odds-ratio scores calculated over the documents acquired so far was one of the most consistently accurate query-generation methods. We also describe experiments using a handful of words elicited from a user instead of initial documents and show that the methods perform similarly. Experiments applying the same approach to multiple languages are also presented showing that our approach generalizes to a variety of languages. 1.

Rocchio's similarity-based Relevance feedback algorithm, one of the most important query reformation methods in information retrieval, is essentially an adaptive supervised learning algorithm from examples.

This paper addresses the problem of specifying, retrieving, filtering and rating Web searches so as to improve the relevance and quality of hits, based on the user’s search intent and preferences. We present a methodology and architecture for an agent-based system, called WebSifter II, that captures the semantics of a user’s decisionoriented search intent, transforms the semantic query into target queries for existing search engines, and then ranks the resulting page hits according to a user-specified weighted-rating scheme. Users create personalized search taxonomies via our Weighted Semantic-Taxonomy Tree. The terms in the tree can be refined by consulting a web taxonomy agent such as Wordnet. The concepts represented in the tree are then transformed into a collection of queries processed by existing search engines. Each returned page is rated according to userspecified preferences such as semantic relevance, syntactic relevance, categorical match, page popularity and authority/hub rating. 1.

In this article we report our research on building FEA-TURES—an intelligent web search engine that is able to perform real-time adaptive feature (i.e., keyword) and document learning. Not only does FEATURES learn from the user’s document relevance feedback, but it also automatically extracts and suggests indexing keywords relevant to a search query and learns from the user’s keyword relevance feedback so that it is able to speed up its search process and to enhance its search performance. We design two efficient and mutual-benefiting learning algorithms that work concurrently, one for feature learning and the other for document learning. FEA-TURES employs these algorithms together with an internal index database and a real-time meta-searcher to perform adaptive real-time learning to find desired documents with as little relevance feedback from the user as possible. The architecture and performance of FEATURES are also discussed. 1.

In contrast to the adoption of linear additive query updating techniques in existing popular algorithms for user preference retrieval, in this paper we design two types of algorithms, the multiplicative adaptive query expansion algorithm MA and the multiplicative adaptive gradient search algorithm MG, both of which use multiplicative query expansion strategies to adaptively improve the query vector. We prove that algorithm MA has a substantially better mistake bound than the Rocchio's and the Perceptron algorithms in learning a user preference relation determined by a linear classier with a small number of non-zero coecients over the real-valued vector space [0; 1] n . We also show that algorithm MG boosts the usefulness of an index term exponentially, while the gradient descent procedure does so linearly. Our work also generalize the algorithm Winnow in the following aspects: various updating functions may be used; multiplicative updating for a weight is dependent on the value of the corresponding index term, which is more realistic and applicable to real-valued vector space; and nally, a number of documents which may or may not be counterexamples to the algorithm's current classication are allowed. Practical implementations of algorithms MA and MG have been underway in the next stage development of our intelligent web search tools. 1 Vector Space and User Preference Let R be the set of all real values, and let R + be the set of all non-negative real values. Let n be a positive integer. In the vector space model in information retrieval [11], a collection of n indexing terms T 1 ; T 2 ; : : : ; Tn are used to represent documents and queries. Each document d is represented as a vector d = (d 1 ; : : : ; dn ) such that for any i, 1 i n, the i-th component ...

In this paper we report our research on building FEATURES - an intelligent web search engine that is able to perform real-time adaptive feature (i.e., keyword) and document learning. Not only does FEATURES learn from the user's document relevance feedback, but also automatically extracts and suggests indexing keywords relevant to a search query and learns from the user's keyword relevance feedback so that it is able to speed up its search process and to enhance its search performance. We design two efficient and mutual-benefiting learning algorithms that work concurrently, one for feature learning and the other for document learning. FEATURES employs these algorithms together with an internal index database and a real-time meta-searcher so to perform adaptive real-time learning to find desired documents with as little relevance feedback from the user as possible. The architecture and performance of FEATURES are also discussed.

The Web is a valuable source of language specific resources but the process of collecting, organizing and utilizing these resources is difficult. We describe CorpusBuilder, an approach for automatically generating Web-search queries for collecting documents in a minority language. It differs from pseudo-relevance feedback in that retrieved documents are labeled by an automatic language classifier as relevant or irrelevant, and this feedback is used to generate new queries. We experiment with various query-generation methods and query-lengths to find inclusion/exclusion terms that are helpful for retrieving documents in the target language and find that using odds-ratio scores calculated over the documents acquired so far was one of the most consistently accurate query-generation methods. We also describe experiments using a handful of words elicited from a user instead of initial documents and show that the methods perform similarly. Experiments applying the same approach to multiple languages are also presented showing that our approach generalizes to a variety of languages. 1.