Minimal Contraction of Preference Relations

AAAI Conferences

Changing preferences is very common in real life. The expressive power of the operations of preference change introduced so far in the literature is limited to adding new information about preference and equivalence. Here, we discuss the operation of discarding preferences: preference contraction. We argue that the property of minimality and the preservation of strict partial orders are crucial for contractions. Contractions can be further constrained by specifying which preferences should not be contracted. We provide algorithms for computing minimal and minimal preference-protecting contraction. We also show some preference query optimization techniques which can be used in the presence of contraction.


Reciprocal Hash Tables for Nearest Neighbor Search

AAAI Conferences

Recent years have witnessed the success of hashingtechniques in approximate nearest neighbor search. Inpractice, multiple hash tables are usually employed toretrieve more desired results from all hit buckets ofeach table. However, there are rare works studying theunified approach to constructing multiple informativehash tables except the widely used random way. In thispaper, we regard the table construction as a selectionproblem over a set of candidate hash functions. Withthe graph representation of the function set, we proposean efficient solution that sequentially applies normal-ized dominant set to finding the most informative andindependent hash functions for each table. To furtherreduce the redundancy between tables, we explore thereciprocal hash tables in a boosting manner, where thehash function graph is updated with high weights em-phasized on the misclassified neighbor pairs of previoushash tables. The construction method is general andcompatible with different types of hashing algorithmsusing different feature spaces and/or parameter settings.Extensive experiments on two large-scale benchmarksdemonstrate that the proposed method outperforms bothnaive construction method and state-of-the-art hashingalgorithms, with up to 65.93% accuracy gains.


SemMemDB: In-Database Knowledge Activation

AAAI Conferences

Semantic networks are a popular way of simulating human memory in ACT-R-like cognitive architectures. However, existing implementations fall short in their ability to efficiently work with very large networks required for full-scale simulations of human memories. In this paper, we present SemMemDB, an in-database realization of semantic networks and spreading activation. We describe a relational representation for semantic networks and an efficient SQL-based spreading activation algorithm. We provide a simple interface for users to invoke retrieval queries. The key benefits of our approach are: (1) Databases have mature query engines and optimizers that generate efficient query plans for memory activation and retrieval; (2) Databases can provide massive storage capacity to potentially support human-scale memories; (3) Spreading activation is implemented in SQL, a widely-used query language for big data analytics. We evaluate SemMemDB in a comprehensive experimental study using DBPedia, a web-scale ontology constructed from the Wikipedia corpus. The results show that our system runs over 500 times faster than previous works.


Planning to guide concept understanding in the WWW

AAAI Conferences

It also has the ability to generate operators during planning from Web pages using keyword extraction methods. When a user wants to understand a concept, it is useful to browse for relevant Web pages in the WW'W. However, in general, this task is very hard because the user does not know where such Web pages are, and has to search for them in the vast WWW search space.


Integrating Planning and Execution for Information Gathering

AAAI Conferences

The task of information gathering requires locating, retrieving, and integrating information from large numbers of distributed and heterogeneous information sources. In this environment flexibility and efficiency are critical. The usual approach to generating a plan for processing information and then executing it is inflexible and may be very inefficient if problems arise in the query processing. The problem is that actions may fail, the system has incomplete information about the environment, new goals may arise at any time, and the overall efficiency is important.