In this paper we present a novel method, Knowledge Persistence (), for faster evaluation of Knowledge Graph (KG) completion approaches. Current ranking-based evaluation is quadratic in the size of the KG, leading to long evaluation times and consequently a high carbon footprint. addresses this by representing the topology of the KG completion methods through the lens of topological data analysis, concretely using persistent homology. The characteristics of persistent homology allow to evaluate the quality of the KG completion looking only at a fraction of the data. Experimental results on standard datasets show that the proposed metric is highly correlated with ranking metrics (Hits@N, MR, MRR). Performance evaluation shows that is computationally efficient: In some cases, the evaluation time (validation+test) of a KG completion method has been reduced from 18 hours (using Hits@10) to 27 seconds (using ), and on average (across methods & data) reduces the evaluation time (validation+test) by ≈ 99.96%.

We present a novel semantic context prior-based venue recommendation system that uses only the title and the abstract of a paper. Based on the intuition that the text in the title and abstract have both semantic and syntactic components, we demonstrate that a joint training of a semantic feature extractor and syntactic feature extractor collaboratively leverages meaningful information that helps to provide venues for papers. The proposed methodology that we call DeSCoVeR at first elicits these semantic and syntactic features using a Neural Topic Model and text classifier respectively. The model then executes a transfer learning optimization procedure to perform a contextual transfer between the feature distributions of the Neural Topic Model and the text classifier during the training phase. DeSCoVeR also mitigates the document-level label bias using a Causal back-door path criterion and a sentence-level keyword bias removal technique. Experiments on the DBLP dataset show that DeSCoVeR outperforms the state-of-the-art methods.

We examine the applicability of Composite Items (CIs) for generating customized travel packages consisting of Points of Interest (POIs) in a given city. CIs have been shown to serve complex information needs such as selecting books for a reading club, identifying a set of products for a promotion, or planning a city tour. In the travel domain, a synthesized view of travel options in a city can be provided with a set of cohesive CIs, each of which is covering a different region in the city. In this paper, we attempt to understand the benefit of letting users customize travel packages, and examine the relationship between customization and personalization. For personalization, we gather user preferences on POI features when available or on latent topics extracted from POI tags. For customization, we develop a framework within which a user interacts with proposed travel packages and the system suggests new CIs according to refined user preferences. Our experiments reveal a tension between personalization and the cohesiveness of items forming each CI. As a result, customization is necessary to find a balance between POI personalization and CI cohesiveness. We also show that the refined user preferences obtained from customization in one city help build better travel packages in another city.

A traditional relational database can evaluate complex queries \nbut requires users to precisely express their information need. \nBut users often do not know what information is available \nin a database, and hence cannot correctly express their information need. Traditional databases do not provide convenient means for users to gain familiarity with the data. \nIn this paper, we study the problem of exploratory search, \nwhich a user may wish to perform to get an understanding of \nthe data set. We note that users often have some decisions \nalready made, so what they need is not an overall database \nsummary, but rather a summary “in context” of the relevant portion of the database. Towards this end, we devise a \nnovel data summarization technique called the Conditional \nAttribute Dependency (CAD) View, which shows the conditional dependencies between attribute values conditioned on \napplied selections. The CAD View can help users to gain familiarity with structured datasets in an attribute-wise manner. \nTo evaluate the CAD View, we perform a user study comprising three complex exploratory tasks on a real dataset. \nOur studies show that users are able to do all the tasks \nabout 4-5 times faster and with better accuracy using the \nCAD View compared to the data summary shown in faceted \nnavigation, which is currently the most popular search interface for e-commerce and has support for exploratory search.

Numerous applications such as wireless communication and telematics need to keep track of evolution of spatio-temporal data for a limited past. Limited retention may even be required by regulations. In general, each data entry can have its own user specified lifetime. It is desired that expired entries are automatically removed by the system through some garbage collection mechanism. This kind of limited retention can be achieved by using a sliding window semantics similar to that from stream data processing. However, due to the large volume and relatively long lifetime of data in the aforementioned applications (in contrast to the real-time transient streaming data), the sliding window here needs to be maintained for data on disk rather than in memory. It is a new challenge to provide fast access to the information from the recent past and, at the same time, facilitate efficient deletion of the expired entries. In this paper, we propose a disk based, two-layered, sliding window indexing scheme for discretely moving spatio-temporal data. Our index can support efficient processing of standard time slice and interval queries and delete expired entries with almost no overhead. In existing historical spatio-temporal indexing techniques, deletion is either infeasible or very inefficient. Our sliding window based processing model can support both current and past entries, while many existing historical spatio-temporal indexing techniques cannot keep these two types of data together in the same index. Our experimental comparison with the best known historical index (i.e., the MV3R tree) for discretely moving spatio-temporal data shows that our index is about five times faster in terms of insertion time and comparable in terms of search performance. MV3R follows a partial persistency model, whereas our index can support very efficient deletion and update.

A relational database often yields a large set of tuples as the result of a query. Users browse this result set to find the information they require. If the result set is large, there may be many pages of data to browse. Since results comprise tuples of alphanumeric values that have few visual markers, it is hard to browse the data quickly, even if it is sorted.