Syntactic search relies on keywords contained in a query to find suitable
documents. So, documents that do not contain the keywords but contain
information related to the query are not retrieved. Spreading activation is an
algorithm for finding latent information in a query by exploiting relations
between nodes in an associative network or semantic network. However, the
classical spreading activation algorithm uses all relations of a node in the
network that will add unsuitable information into the query. In this paper, we
propose a novel approach for semantic text search, called
query-oriented-constrained spreading activation that only uses relations
relating to the content of the query to find really related information.
Experiments on a benchmark dataset show that, in terms of the MAP measure, our
search engine is 18.9% and 43.8% respectively better than the syntactic search
and the search using the classical constrained spreading activation.
KEYWORDS: Information Retrieval, Ontology, Semantic Search, Spreading
Activation
Authors' comments: 12pages, will be published in The International Journal of Artificial
Intelligence & Applications (IJAIA). arXiv admin note: text overlap with
arXiv:1807.07967
Conventional remote sensing image retrieval (RSIR) systems usually perform
single-label retrieval where each image is annotated by a single label
representing the most significant semantic content of the image. This
assumption, however, ignores the complexity of remote sensing images, where an
image might have multiple classes (i.e., multiple labels), thus resulting in
worse retrieval performance. We therefore propose a novel multi-label RSIR
approach with fully convolutional networks (FCN). In our approach, we first
train a FCN model using a pixel-wise labeled dataset,and the trained FCN is
then used to predict the segmentation maps of each image in the considered
archive. We finally extract region convolutional features of each image based
on its segmentation map.The region features can be either used to perform
region-based retrieval or further post-processed to obtain a feature vector for
similarity measure. The experimental results show that our approach achieves
state-of-the-art performance in contrast to conventional single-label and
recent multi-label RSIR approaches.
Authors' comments: 8 pages
Named entities (NE) are objects that are referred to by names such as people,
organizations and locations. Named entities and keywords are important to the
meaning of a document. We propose a generalized vector space model that
combines named entities and keywords. In the model, we take into account
different ontological features of named entities, namely, aliases, classes and
identifiers. Moreover, we use entity classes to represent the latent
information of interrogative words in Wh-queries, which are ignored in
traditional keyword-based searching. We have implemented and tested the
proposed model on a TREC dataset, as presented and discussed in the paper.
Authors' comments: 5 pages, in Vietnamese. information retrieval, vector space model,
ontology, named entity, keyword. Accepted by Vietnamese Journal on
Information Technologies and Communications
Named entities have been considered and combined with keywords to enhance
information retrieval performance. However, there is not yet a formal and
complete model that takes into account entity names, classes, and identifiers
together. Our work explores various adaptations of the traditional Vector Space
Model that combine different ontological features with keywords, and in
different ways. It shows better performance of the proposed models as compared
to the keyword-based Lucene, and their advantages for both text retrieval and
representation of documents and queries.
Authors' comments: 10 pages, will be in PRICAI. arXiv admin note: substantial text
overlap with arXiv:1807.05576
The popular approaches to recommendation and ad-hoc retrieval tasks are
largely distinct in the literature. In this work, we argue that many
recommendation problems can also be cast as ad-hoc retrieval tasks. To
demonstrate this, we build a solution for the RecSys 2018 Spotify challenge by
combining standard ad-hoc retrieval models and using popular retrieval tools
sets. We draw a parallel between the playlist continuation task and the task of
finding good expansion terms for queries in ad-hoc retrieval, and show that
standard pseudo-relevance feedback can be effective as a collaborative
filtering approach. We also use ad-hoc retrieval for content-based
recommendation by treating the input playlist title as a query and associating
all candidate tracks with meta-descriptions extracted from the background data.
The recommendations from these two approaches are further supplemented by a
nearest neighbor search based on track embeddings learned by a popular neural
model. Our final ranked list of recommendations is produced by a learning to
rank model. Our proposed solution using ad-hoc retrieval models achieved a
competitive performance on the music recommendation task at RecSys 2018
challenge---finishing at rank 7 out of 112 participating teams and at rank 5
out of 31 teams for the main and the creative tracks, respectively.
We consider the phase retrieval problem for signals that belong to a union of
subspaces. We assume that amplitude measurements of the signal of length $n$
are observed after passing it through a random $m \times n$ measurement matrix.
We also assume that the signal belongs to the span of a single $d$-dimensional
subspace out of $R$ subspaces, where $d\ll n$. We assume the knowledge of all
possible subspaces, but the true subspace of the signal is unknown. We present
an algorithm that jointly estimates the phase of the measurements and the
subspace support of the signal. We discuss theoretical guarantees on the
recovery of signals and present simulation results to demonstrate the empirical
performance of our proposed algorithm. Our main result suggests that if
properly initialized, then $O(d+\log R)$ random measurements are sufficient for
phase retrieval if the unknown signal belongs to the union of $R$
low-dimensional subspaces.
Authors' comments: working draft
We aim to find a solution $\bm{x}\in\mathbb{C}^n$ to a system of quadratic
equations of the form $b_i=\lvert\bm{a}_i^*\bm{x}\rvert^2$, $i=1,2,\ldots,m$,
e.g., the well-known NP-hard phase retrieval problem. As opposed to recently
proposed state-of-the-art nonconvex methods, we revert to the semidefinite
relaxation (SDR) PhaseLift convex formulation and propose a successive and
incremental nonconvex optimization algorithm, termed as \texttt{IncrePR}, to
indirectly minimize the resulting convex problem on the cone of positive
semidefinite matrices. Our proposed method overcomes the excessive
computational cost of typical SDP solvers as well as the need of a good
initialization for typical nonconvex methods. For Gaussian measurements, which
is usually needed for provable convergence of nonconvex methods,
\texttt{IncrePR} with restart strategy outperforms state-of-the-art nonconvex
solvers with a sharper phase transition of perfect recovery and typical convex
solvers in terms of computational cost and storage. For more challenging
structured (non-Gaussian) measurements often occurred in real applications,
such as transmission matrix and oversampling Fourier transform,
\texttt{IncrePR} with several restarts can be used to find a good initial
guess. With further refinement by local nonconvex solvers, one can achieve a
better solution than that obtained by applying nonconvex solvers directly when
the number of measurements is relatively small. Extensive numerical tests are
performed to demonstrate the effectiveness of the proposed method.
Authors' comments: 20 pages, 25 figures
Text search based on lexical matching of keywords is not satisfactory due to
polysemous and synonymous words. Semantic search that exploits word meanings,
in general, improves search performance. In this paper, we survey WordNet-based
information retrieval systems, which employ a word sense disambiguation method
to process queries and documents. The problem is that in many cases a word has
more than one possible direct sense, and picking only one of them may give a
wrong sense for the word. Moreover, the previous systems use only word forms to
represent word senses and their hypernyms. We propose a novel approach that
uses the most specific common hypernym of the remaining undisambiguated
multi-senses of a word, as well as combined WordNet features to represent word
meanings. Experiments on a benchmark dataset show that, in terms of the MAP
measure, our search engine is 17.7% better than the lexical search, and at
least 9.4% better than all surveyed search systems using WordNet.
Keywords Ontology, word sense disambiguation, semantic annotation, semantic
search.
Authors' comments: 6pages, Will be in proceedings of the 5th International Conference on
Intelligent Computing and Information Systems (ICICIS-2011), in cooperation
with ACM. 30 June to 3 July, 2011, Cairo, Egypt
In many retrieval problems, where we must retrieve one or more entries from a
gallery in response to a probe, it is common practice to learn to do by
directly comparing the probe and gallery entries to one another. In many
situations the gallery and probe have common covariates -- external variables
that are common to both. In principle it is possible to perform the retrieval
based merely on these covariates. The process, however, becomes gated by our
ability to recognize the covariates for the probe and gallery entries
correctly.
In this paper we analyze optimal strategies for retrieval based only on
matching covariates, when the recognition of the covariates is itself
inaccurate. We investigate multiple problems: recovering one item from a
gallery of $N$ entries, matching pairs of instances, and retrieval from large
collections. We verify our analytical formulae through experiments to verify
their correctness in practical settings.
Authors' comments: support material for "Disjoint Mapping Network for Cross-modal
Matching of Voices and Faces"
During maintenance, software developers deal with numerous change requests
that are written in an unstructured fashion using natural language. Such
natural language texts illustrate the change requirement involving various
domain related concepts. Software developers need to find appropriate search
terms from those concepts so that they could locate the possible locations in
the source code using a search technique. Once such locations are identified,
they can implement the requested changes there. Studies suggest that developers
often perform poorly in coming up with good search terms for a change task. In
this paper, we propose a novel technique--STRICT--that automatically identifies
suitable search terms for a software change task by analyzing its task
description using two information retrieval (IR) techniques-- TextRank and
POSRank. These IR techniques determine a term's importance based on not only
its co-occurrences with other important terms but also its syntactic
relationships with them. Experiments using 1,939 change requests from eight
subject systems report that STRICT can identify better quality search terms
than baseline terms from 52%--62% of the requests with 30%--57% Top-10
retrieval accuracy which are promising. Comparison with two state-of-the-art
techniques not only validates our empirical findings and but also demonstrates
the superiority of our technique.
Authors' comments: The 24th IEEE International Conference on Software Analysis,
Evolution, and Reengineering (SANER 2017), pp. 79--90, Klagenfurt, Austria,
February 2017
Vector-Quantized Variational Autoencoders (VQ-VAE)[1] provide an unsupervised
model for learning discrete representations by combining vector quantization
and autoencoders. In this paper, we study the use of VQ-VAE for representation
learning for downstream tasks, such as image retrieval. We first describe the
VQ-VAE in the context of an information-theoretic framework. We show that the
regularization term on the learned representation is determined by the size of
the embedded codebook before the training and it affects the generalization
ability of the model. As a result, we introduce a hyperparameter to balance the
strength of the vector quantizer and the reconstruction error. By tuning the
hyperparameter, the embedded bottleneck quantizer is used as a regularizer that
forces the output of the encoder to share a constrained coding space such that
learned latent features preserve the similarity relations of the data space. In
addition, we provide a search range for finding the best hyperparameter.
Finally, we incorporate the product quantization into the bottleneck stage of
VQ-VAE and propose an end-to-end unsupervised learning model for the image
retrieval task. The product quantizer has the advantage of generating
large-size codebooks. Fast retrieval can be achieved by using the lookup tables
that store the distance between any pair of sub-codewords. State-of-the-art
retrieval results are achieved by the learned codebooks.
In this paper, we study jointly query reformulation and document relevance
estimation, the two essential aspects of information retrieval (IR). Their
interactions are modelled as a two-player strategic game: one player, a query
formulator, taking actions to produce the optimal query, is expected to
maximize its own utility with respect to the relevance estimation of documents
produced by the other player, a retrieval modeler; simultaneously, the
retrieval modeler, taking actions to produce the document relevance scores,
needs to optimize its likelihood from the training data with respect to the
refined query produced by the query formulator. Their equilibrium or equilibria
will be reached when both are the best responses to each other. We derive our
equilibrium theory of IR using normal-form representations: when a standard
relevance feedback algorithm is coupled with a retrieval model, they would
share the same objective function and thus form a partnership game; by
contrast, pseudo relevance feedback pursues a rather different objective than
that of retrieval models, therefore the interaction between them would lead to
a general-sum game (though implicitly collaborative). Our game-theoretical
analyses not only yield useful insights into the two major aspects of IR, but
also offer new practical algorithms for achieving the equilibrium state of
retrieval which have been shown to bring consistent performance improvements in
both text retrieval and item recommendation.
Authors' comments: Accepted in ICTIR 2018
Imaging systems' performance at low light intensity is affected by shot
noise, which becomes increasingly strong as the power of the light source
decreases. In this paper we experimentally demonstrate the use of deep neural
networks to recover objects illuminated with weak light and demonstrate better
performance than with the classical Gerchberg-Saxton phase retrieval algorithm
for equivalent signal over noise ratio. Prior knowledge about the object is
implicitly contained in the training data set and feature detection is possible
for a signal over noise ratio close to one. We apply this principle to a phase
retrieval problem and show successful recovery of the object's most salient
features with as little as one photon per detector pixel on average in the
illumination beam. We also show that the phase reconstruction is significantly
improved by training the neural network with an initial estimate of the object,
as opposed as training it with the raw intensity measurement.
Authors' comments: 8 pages, 5 figures
The interpretation of the experimental data collected by testing systems
across input datasets and model parameters is of strategic importance for
system design and implementation. In particular, finding relationships between
variables and detecting the latent variables affecting retrieval performance
can provide designers, engineers and experimenters with useful if not necessary
information about how a system is performing. This paper discusses the use of
Structural Equation Modelling (SEM) in providing an in-depth explanation of
evaluation results and an explanation of failures and successes of a system; in
particular, we focus on the case of Information Retrieval.
Most image retrieval methods use global features that aggregate local
distinctive patterns into a single representation. However, the aggregation
process destroys the relative spatial information by considering orderless sets
of local descriptors. We propose to integrate relative spatial information into
the aggregation process by taking into account co-occurrences of local patterns
in a tensor framework. The resulting signature called Improved Spatial Tensor
Aggregation (ISTA) is able to reach state of the art performances on well known
datasets such as Holidays, Oxford5k and Paris6k.
Authors' comments: 8 pages, 2 figures and 1 table. Draft paper for conference, IEEE
International Conference on Image Processing (ICIP) 2018
The landmark recognition problem is far from being solved, but with the use
of features extracted from intermediate layers of Convolutional Neural Networks
(CNNs), excellent results have been obtained. In this work, we propose some
improvements on the creation of R-MAC descriptors in order to make the
newly-proposed R-MAC+ descriptors more representative than the previous ones.
However, the main contribution of this paper is a novel retrieval technique,
that exploits the fine representativeness of the MAC descriptors of the
database images. Using this descriptors called "db regions" during the
retrieval stage, the performance is greatly improved. The proposed method is
tested on different public datasets: Oxford5k, Paris6k and Holidays. It
outperforms the state-of-the- art results on Holidays and reached excellent
results on Oxford5k and Paris6k, overcame only by approaches based on
fine-tuning strategies.
In this short note, we consider the worst case noise robustness of any phase
retrieval algorithm which aims to reconstruct all nonvanishing vectors
$\mathbf{x} \in \mathbb{C}^d$ (up to a single global phase multiple) from the
magnitudes of an arbitrary collection of local correlation measurements.
Examples of such measurements include both spectrogram measurements of
$\mathbf{x}$ using locally supported windows and masked Fourier transform
intensity measurements of $\mathbf{x}$ using bandlimited masks. As a result,
the robustness results considered herein apply to a wide range of both
ptychographic and Fourier ptychographic imaging scenarios. In particular, the
main results imply that the accurate recovery of high-resolution images of
extremely large samples using highly localized probes is likely to require an
extremely large number of measurements in order to be robust to worst case
measurement noise, independent of the recovery algorithm employed. Furthermore,
recent pushes to achieve high-speed and high-resolution ptychographic imaging
of integrated circuits for process verification and failure analysis will
likely need to carefully balance probe design (e.g., their effective
time-frequency support) against the total number of measurements acquired in
order for their imaging techniques to be stable to measurement noise, no matter
what reconstruction algorithms are applied.
Recently, hashing techniques have gained importance in large-scale retrieval
tasks because of their retrieval speed. Most of the existing cross-view
frameworks assume that data are well paired. However, the fully-paired
multiview situation is not universal in real applications. The aim of the
method proposed in this paper is to learn the hashing function for semi-paired
cross-view retrieval tasks. To utilize the label information of partial data,
we propose a semi-supervised hashing learning framework which jointly performs
feature extraction and classifier learning. The experimental results on two
datasets show that our method outperforms several state-of-the-art methods in
terms of retrieval accuracy.
Authors' comments: 6 pages, 5 figures, 2 tables
We consider a game-theoretic model of information retrieval with strategic
authors. We examine two different utility schemes: authors who aim at
maximizing exposure and authors who want to maximize active selection of their
content (i.e. the number of clicks). We introduce the study of author learning
dynamics in such contexts. We prove that under the probability ranking
principle (PRP), which forms the basis of the current state of the art ranking
methods, any better-response learning dynamics converges to a pure Nash
equilibrium. We also show that other ranking methods induce a strategic
environment under which such a convergence may not occur.
Neural network approaches have recently shown to be effective in several
information retrieval (IR) tasks. However, neural approaches often require
large volumes of training data to perform effectively, which is not always
available. To mitigate the shortage of labeled data, training neural IR models
with weak supervision has been recently proposed and received considerable
attention in the literature. In weak supervision, an existing model
automatically generates labels for a large set of unlabeled data, and a machine
learning model is further trained on the generated "weak" data. Surprisingly,
it has been shown in prior art that the trained neural model can outperform the
weak labeler by a significant margin. Although these obtained improvements have
been intuitively justified in previous work, the literature still lacks
theoretical justification for the observed empirical findings. In this position
paper, we propose to theoretically study weak supervision, in particular for IR
tasks, e.g., learning to rank. We briefly review a set of our recent
theoretical findings that shed light on learning from weakly supervised data,
and provide guidelines on how train learning to rank models with weak
supervision.
Authors' comments: A position paper accepted to the 2018 ACM SIGIR Workshop on Learning
from Limited or Noisy Data for Information Retrieval (LND4IR)
