In keyword spotting from handwritten documents, the word similarity is usually computed by combining character similarities. Converting similarity to probabilistic confidence is beneficial for context fusion and threshold selection. In this paper, we propose to directly estimate the posterior probability of candidate
support context–keyword queries. Our calculations indicate that the technique yields improvements in the average query hop count while reducing the amount of state stored on each node. The use of Preference Lists can further reduce the average hop count through bypassing previously traversed segments of the structure.
depend probabilistically both on other properties of that object and on properties of related objects. In this paper an attempt is made to heed keywords extraction. The keywords are not only essential for academic papers but also important for web page retrieval, text mining, and document classification. In this paper, a C
corpus. Using a bigram phoneme language model, phoneme recognition experiments are performed on a two hour independent test set using the Viterbi decoding which show a relative 33.3% improvement by our CD-DNN acoustic model. We then present a filler based Hybrid DNN-HMM Keyword Spotting KWS system which to our knowledge is
For text-query-based keyword spotting from handwritten Chinese documents, the index is usually organized as a candidate lattice to overcome the ambiguity of character segmentation. Each edge in the lattice denotes a candidate character associated with a candidate class. Character similarity (between character and
This paper proposes a lattice-based method for keyword spotting in online Chinese handwriting to improve the trade-off between accuracy and speed, and to overcome the out-of-vocabulary (OOV) problem of lexicon-driven approach. Using a character string recognition algorithm, the lattice-based method generates a
This paper proposes a method for keyword spotting in offline Chinese handwritten documents using a statistical model. On a text query word, the method measures the similarity between the query word and every candidate word in the document by combining a character classifier and four classifiers characterizing the
connect any two keywords, (2) The eccentricity of keyword vertices, a well known path measure. Our analysis shows that K-H networks conform to the phenomenon of the shrinking world. Specifically, it shows that the number of vertices of any two keywords, that were not originally connected in the K-K networks, is exactly three
The paper deals with the development of acoustic keyword spotter (KWS) meeting requirements of a real user from the security community. While the basic scheme of the KWS is relatively standard, it uses novel features derived by a hierarchy of neural networks, and score normalization trained to maximize a user-like
We propose a max-pooling based loss function for training Long Short-Term Memory (LSTM) networks for small-footprint keyword spotting (KWS), with low CPU, memory, and latency requirements. The max-pooling loss training can be further guided by initializing with a cross-entropy loss trained network. A posterior
Existing methods for Blog keyword extraction usually exploit the context in the specified blog. In this paper, we propose to provide a knowledge context by using small number of nearest neighbor blogs to improve keyword extraction performance. Specifically, knowledge context is build by adding several topic related
Internet sources using a keyword-based place model as input. Based on external relevance criteria the system finds and pre-selects only those sources that are more relevant, and an adaptive scheduling algorithm continuously select content that are relevant, timely, in accordance with the place model, sensitive to immediate
In this paper, we propose a novel image search scheme is contextual image search with keyword input. It is different from conventional image search schemes. it consist of three step process, first one is context extraction to distinguish the image entities of the same name, second step is conceptualization to convert
Textual web pages dominate web search engines nowadays. However, there is also a striking increase of structured data on the web. Efficient keyword query processing on structured data has attracted enough attention, but effective query understanding has yet to be investigated. In this paper, we focus on the problem of
by the network -- descriptive keywords, or tags. In this paper we present a model that enables keyword discovery methods through the interpretation of the network as a graph, solely relying on keywords that categorize or describe productive items. The model and keyword discovery methods presented in this paper avoid
expressed in terms of keywords, over several XML streams. However, there are few algorithms that evaluate this kind of query. One of them is MKStream, which is the current state-of-the-art algorithm for processing keyword-based queries over XML streams. In order to improve scalability, in this paper we introduce PMKStream
developed by implementing the keyword stripping using the Porter Stemmer algorithm. This could make the keyword search more efficient, as the root or stem word is only considered. Experimental results on two public spam corpuses are also discussed at the end.
This paper presents an objective keyword selection method called visualness with Lesk disambiguation (VLD) for describing educational videos with semantic tags. It extends the work on automatically extracting and associating meaningful keywords carried out in ‘semantic tags for lecture videos’ for
Choosing descriptive keywords to best describe digital media content is crucial for many applications, especially those involving content-based indexing or retrieval. Traditionally such keywords are selected manually, which is labor intensive, restrictive to a limited set of words and inherently subjective to the
In the last three decades, engineering education research (EER) has made remarkable progress towards a field of interdisciplinary scholarship. This paper defines EER by developing a keyword-based scheme for exploring EER-related scientific publications and collaboration. The keyword-based scheme refers to a conceptual
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.