Skip to main content

Xiaodan Zhang

Drexel University, College of Information Science and Technology, Graduate Student

Followers

10

Following

8

Public Views

Interests

Uploads

Papers by Xiaodan Zhang

Context-sensitive semantic smoothing for the language modeling approach to genomic IR

Proceedings of the 29th annual international ACM SIGIR conference on Research and development in information retrieval, 2006

Semantic smoothing, which incorporates synonym and sense information into the language models, is... more Semantic smoothing, which incorporates synonym and sense information into the language models, is effective and potentially significant to improve retrieval performance. The implemented semantic smoothing models, such as the translation model which statistically maps document terms to query terms, and a number of works that have followed have shown good experimental results. However, these models are unable to incorporate contextual information. Thus, the resulting translation might be mixed and fairly general. To overcome this limitation, we propose a novel context-sensitive semantic smoothing method that decomposes a document or a query into a set of weighted context-sensitive topic signatures and then translate those topic signatures into query terms. In detail, we solve this problem through (1) choosing concept pairs as topic signatures and adopting an ontology-based approach to extract concept pairs; (2) estimating the translation model for each topic signature using the EM algorithm; and (3) expanding document and query models based on topic signature translations. The new smoothing method is evaluated on TREC 2004/05 Genomics Track collections and significant improvements are obtained. The MAP (mean average precision) achieves a 33.6% maximal gain over the simple language model, as well as a 7.8% gain over the language model with context-insensitive semantic smoothing.

A Semantic Approach for Mining Hidden Links from Complementary and Non-interactive Biomedical Literature

Proceedings of the 2006 SIAM International Conference on Data Mining, 2006

Two complementary and non-interactive literature sets of articles, when they are considered toget... more Two complementary and non-interactive literature sets of articles, when they are considered together, can reveal useful information of scientific interest not apparent in either of the two sets alone. Swanson called the existence of such hidden links as undiscovered public knowledge (UPK). The novel connection between Raynaud disease and fish oils was uncovered from complementary and non-interactive biomedical literature by Swanson in 1986. Since then, there have been many approaches to uncover UPK by mining the biomedical literature. These earlier works, however, required substantial manual intervention to reduce the number of possible connections. This paper proposes a semantic-based mining model for undiscovered public knowledge using the biomedical literature. Our method replaces manual ad-hoc pruning by using semantic knowledge from t he biomedical ontologies. Using the semantic types and semantic relationships of the biomedical concepts, our prototype system can identify the relevant concepts collected from Medline and generate the novel hypothesis between these concepts. The system successfully replicates Swanson's two famous discoveries: Raynaud disease/fish oils and migraine/magnesium. Compared with previous approaches such as LSI-based and traditional association rule-based methods, our method generates much fewer but more relevant novel hypotheses, and requires much less human intervention in the discovery procedure

Semantic Smoothing for Bayesian Text Classification with Small Training Data

Proceedings of the 2008 SIAM International Conference on Data Mining, 2008

Bayesian text classifiers face a common issue which is referred to as data sparsity problem, espe... more Bayesian text classifiers face a common issue which is referred to as data sparsity problem, especially when the size of training data is very small. The frequently used Laplacian smoothing and corpus-based background smoothing are not effective in handling it. Instead, we propose a novel semantic smoothing method to address the sparse problem. Our method extracts explicit topic signatures (e.g. words, multiword phrases, and ontologybased concepts) from a document and then statistically maps them into single-word features. We conduct comprehensive experiments on three testing collections (OHSUMED, LATimes, and 20NG) to compare semantic smoothing with other approaches. When the size of training documents is small, the bayesian classifier with semantic smoothing not only outperforms the classifiers with background smoothing and Laplacian smoothing, but also beats the state-of-the-art active learning classifiers and SVM classifiers. In this paper, we also compare three types of topic signatures with respect to their effectiveness and efficiency for semantic smoothing.

A Comparative Study of Ontology Based Term Similarity Measures on PubMed Document Clustering

Lecture Notes in Computer Science

Recent research shows that ontology as background knowledge can improve document clustering quali... more Recent research shows that ontology as background knowledge can improve document clustering quality with its concept hierarchy knowledge. Previous studies take term semantic similarity as an important measure to incorporate domain knowledge into clustering process such as clustering initialization and term re-weighting. However, not many studies have been focused on how different types of term similarity measures affect the clustering performance for a certain domain. In this paper, we conduct a comparative study on how different semantic similarity measures of term including path based similarity measure, information content based similarity measure and feature based similarity measure affect document clustering. We evaluate term re-weighting as an important method to integrate domain ontology to clustering process. Meanwhile, we apply k-means clustering on one real-world text dataset, our own corpus generated from PubMed. Experiment results on 8 different semantic measures have shown that: (1) there is no a certain type of similarity measures that significantly outperforms the others; (2) Several similarity measures have rather more stable performance than the others; (3) term re-weighting has positive effects on medical document clustering, but might not be significant when documents are short of terms.

Integration of cluster ensemble and EM based text mining for microarray gene cluster identification and annotation

Proceedings of the 15th ACM international conference on Information and knowledge management - CIKM '06, 2006

In this paper, we design and develop a unified system GE-Miner (Gene Expression Miner) to integra... more In this paper, we design and develop a unified system GE-Miner (Gene Expression Miner) to integrate cluster ensemble, text clustering and multi document summarization and provide an environment for comprehensive gene expression data analysis. We present a novel cluster ensemble approach to generate high quality gene cluster. In our text summarization module, given a gene cluster, our Expectation Maximization (EM) based algorithm can automatically identify subtopics and extract most probable terms for each topic. Then, the extracted top k topical terms from each subtopic are combined to form the biological explanation of each gene cluster. Experimental results demonstrate that our system can obtain high quality clusters and provide informative key terms for the gene clusters.

Relation-Based Document Retrieval for Biomedical IR

Lecture Notes in Computer Science, 2006

In this paper, we explore the direct use of relations in information re- trieval for precision-fo... more In this paper, we explore the direct use of relations in information re- trieval for precision-focused biomedical literature search. A relation is defined as a pair of two concepts which are semantically and syntactically related to each other. Unlike the traditional term-based IR models, our model represents a document by a set of controlled concepts and their binary relations. Since

A Language Modeling Text Mining Approach to the Annotation of Protein Community

Sixth IEEE Symposium on BioInformatics and BioEngineering (BIBE'06), 2006

This paper discusses an ontology based language modeling text mining approach to the annotation o... more This paper discusses an ontology based language modeling text mining approach to the annotation of protein community. Communities appear to play an important role in the functional properties of complex networks. Being able to annotate the identified the community structure in a biological network can help us to understand better the structure and dynamics of biological systems. Traditional method such as Gene Ontology (GO) provides information about the functionality of gene products, but they are not enough to annotate community as for only limited number of proteins in the database, limited protein properties available for annotation and the inability to annotate a group of gene products as a whole. Thus, we present an ontology based mixture language model approach to annotate protein community. Compared to traditional method, we have the following three advantages. First, biomedical literature mining brings much richer information than existed gene databases. Second, the mixture language model can help "purify" the document by eliminating some background noise. Third, using domain ontology, we extract biological concept and concept pairs from abstracts. Biological concept is more meaningful than word or multi-word phrases. Moreover, using concept pairs can deliver much more information and serve as evidence of annotation results. We test our approach on four communities SAGA-SRB, CCR-NOT, RFC and ARP2/3, detected from dataset of interactions for Saccharomyces cerevisae from the General Repository for Interaction Datasets (GRID). Annotation results provide a very coherent indication of functionality of each community. I.

Dragon Toolkit: Incorporating Auto-Learned Semantic Knowledge into Large-Scale Text Retrieval and Mining

19th IEEE International Conference on Tools with Artificial Intelligence(ICTAI 2007), 2007

The majority of text retrieval and mining techniques are still based on exact feature (e.g. words... more The majority of text retrieval and mining techniques are still based on exact feature (e.g. words) matching and unable to incorporate text semantics. Many researchers believe that the extension with semantic knowledge could improve the results and various methods (most of them are heuristic) have been proposed to account for concept hierarchy, synonymy, and other semantic relationships. However, the results with such semantic extension have been mixed, ranging from slight improvements to decreases in effectiveness, mostly likely due to the lack of a formal framework. Instead, we propose a novel method to address the semantic extension within the framework of language modeling. Our method extracts explicit topic signatures from documents and then statistically maps them into singleword features. The incorporation of semantic knowledge then reduces to the smoothing of unigram language models using semantic knowledge. The dragon toolkit reflects our method and its effectiveness is demonstrated by three tasks, text retrieval, text classification, and text clustering.

A semantic-based approach for mining undiscovered public knowledge from biomedical literature

2005 IEEE International Conference on Granular Computing, 2005

The problem of mining undiscovered public knowledge from biomedical literature was exemplified by... more The problem of mining undiscovered public knowledge from biomedical literature was exemplified by Swanson's pioneering work on Raynaud disease/fish-oil discovery in 1986. Since then, there have been many approaches to mine undiscovered public knowledge from biomedical literature. This paper presents a semantic-based approach for mining undiscovered public knowledge from bio-medical literature. The method takes advantages of the biomedical ontologies, MeSH and UMLS, as the source of semantic knowledge. A prototype system Biomedical Semantic-based Knowledge Discovery System (Bio-SbKDS) is designed to uncover novel hypothe-sis/connections hidden in the biomedical literature. Using the semantic types and semantic relations of the bio-medical concepts, Bio-SbKDS can identify the relevant concepts collected from Medline and generate the novel hypothesis between these concepts. Bio-SbKDS successfully replicates Dr. Swanson's two famous discoveries: Raynaud disease/fish oil and migraine/magnesium. Compared with previous approaches, our method searches much less articles, generates much less but more relevant novel hypotheses, requires much less human intervention in the discovery procedure.

Semantic Smoothing for Model-based Document Clustering

Sixth International Conference on Data Mining (ICDM'06), 2006

A document is often full of class-independent "general" words and short of class-specific "core" ... more A document is often full of class-independent "general" words and short of class-specific "core" words, which leads to the difficulty of document clustering. We argue that both problems will be relieved after suitable smoothing of document models in agglomerative approaches and of cluster models in partitional approaches, and hence improve clustering quality. To the best of our knowledge, most modelbased clustering approaches use Laplacian smoothing to prevent zero probability while most similaritybased approaches employ the heuristic TF*IDF scheme to discount the effect of "general" words. Inspired by a series of statistical translation language model for text retrieval, we propose in this paper a novel smoothing method referred to as contextsensitive semantic smoothing for document clustering purpose. The comparative experiment on three datasets shows that semantic smoothing in conjunction with model-based clustering approaches is effective in improving cluster quality.

MaxMatcher: Biological Concept Extraction Using Approximate Dictionary Lookup

Lecture Notes in Computer Science, 2006

Dictionary-based biological concept extraction is still the state-ofthe-art approach to large-sca... more Dictionary-based biological concept extraction is still the state-ofthe-art approach to large-scale biomedical literature annotation and indexing. The exact dictionary lookup is a very simple approach, but always achieves low extraction recall because a biological term often has many variants while a dictionary is impossible to collect all of them. We propose a generic extraction approach, referred to as approximate dictionary lookup, to cope with term variations and implement it as an extraction system called MaxMatcher. The basic idea of this approach is to capture the significant words instead of all words to a particular concept. The new approach dramatically improves the extraction recall while maintaining the precision. In a comparative study on GENIA corpus, the recall of the new approach reaches a 57% recall while the exact dictionary lookup only achieves a 26% recall.

A Comparison of Local Analysis, Global Analysis and Ontology-based Query Expansion Strategies for Bio-medical Literature Search

2006 IEEE International Conference on Systems, Man and Cybernetics, 2006

In this paper we present the design and evaluation of our biomedical literature searching approac... more In this paper we present the design and evaluation of our biomedical literature searching approaches using the TREC 2004 ad hoc retrieval task in the Genomics track. The main approach taken in our system is to expand queries by exploiting the three widely used strategies-local analysis, global analysis, and ontology-based term re-weighting across various search engines. The experimental results show that (1) ontology-based term re-weighting provides the best results among the three query expansion strategies, (2) expanding the initial query with more precise ontology-based term enhances LSI based local analysis substantially, and (3) including context to term re-weighting and LSI further improves the precision. Experimental results also show that the ontology-based term re-weighting with LUCENE or LEMUR search engines increases the average precision by up to 20.3% or 12.1%, respectively, compared to that of the baseline runs. In addition, the LSI-based local analysis increases the average precision by 9.2% with LEMUR search engine. We believe the principles of the term re-weighting strategy and LSI-based local analysis may be extended and utilized in other bio-medical domains.

Semantic representation in text classification using topic signature mapping

2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), 2008

Document representation is one of the crucial components that determine the effectiveness of text... more Document representation is one of the crucial components that determine the effectiveness of text classification tasks. Traditional document representation approaches typically adopt a popular bag-of-word method as the underlying document representation. Although it's a simple and efficient method, the major shortcoming of bag-of-word representation is in the independent of word feature assumption. Many researchers have attempted to address this issue by incorporating semantic information into document representation. In this paper, we study the effect of semantic representation on the effectiveness of text classification systems. We employed a novel semantic smoothing technique to derive semantic information in a form of mapping probability between topic signatures and single-word features. Two classifiers, Naïve Bayes and Support Vector Machine, were selected to carry out the classification experiments. Overall, our topic-signature semantic representation approaches significantly outperformed traditional bag-of-word representation in most datasets. I. INTRODUCTION NE of the major criticisms of bag-of-word document representation is that it treats each individual word as an independent feature, therefore, it ignores any semantic relatedness that might exist between words. Given such assumption, document representation is likely to contain an inaccurate set of features. For example, synonymous words such as automobile and car will be considered as two different features, each with one occurrence, while polysemous words such as bank in financial bank and river bank will be treated as the same feature with two occurrences. Another related issue of bag-of-word representation is data sparseness problem. In many cases, words that appear in training documents of a given class may not appear in testing collection.

Web clustering based on the information of sibling pages

2008 IEEE International Conference on Granular Computing, 2008

This paper is dedicated to investigating the value of information from sibling pages for web page... more This paper is dedicated to investigating the value of information from sibling pages for web page clustering. We use a link-based clustering algorithm to examine the usefulness of sibling links for improving clustering quality. The algorithm is extended by two types of edge weighting techniques. The results of the experiments conducted on WebKB4 dataset prove that: (1) using information from sibling pages can significantly improve clustering quality; (2) sibling pages are more useful than parent and child pages in enhancing clustering performance; (3) weighting and pruning sibling links can not improve the clustering quality. We also conducted an experiment on the citation dataset Cora7. The results indicate that sibling links are not more useful than the direct citation links when used to cluster collections of research papers.

Text Mining the Biomedical Literature for Identification of Potential Virus/Bacterium as Bio-Terrorism Weapons

Integrated Series In Information Systems, 2008

Machine Learning and Knowledge Discovery in Databases

Lecture Notes in Computer Science, 2014

One of the challenges in big data analytics is to efficiently learn and reason collectively about... more One of the challenges in big data analytics is to efficiently learn and reason collectively about extremely large, heterogeneous, incomplete, noisy interlinked data. Collective reasoning requires the ability to exploit both the logical and relational structure in the data and the probabilistic dependencies. In this talk I will overview our recent work on probabilistic soft logic (PSL), a framework for collective, probabilistic reasoning in relational domains. PSL is able to reason holistically about both entity attributes and relationships among the entities. The underlying mathematical framework, which we refer to as a hinge-loss Markov random field, supports extremely efficient, exact inference. This family of graphical models captures logic-like dependencies with convex hinge-loss potentials. I will survey applications of PSL to diverse problems ranging from information extraction to computational social science. Our recent results show that by building on state-of-the-art optimization methods in a distributed implementation, we can solve large-scale problems with millions of random variables orders of magnitude faster than existing approaches. Bio. In 1995, Lise Getoor decided to return to school to get her PhD in Computer Science at Stanford University. She received a National Physical Sciences Consortium fellowship, which in addition to supporting her for six years, supported a summer internship at Xerox PARC, where she worked with Markus Fromherz and his group. Daphne Koller was her PhD advisor; in addition, she worked closely with Nir Friedman, and many other members of the DAGS group, including Avi Pfeffer, Mehran Sahami, Ben Taskar, Carlos Guestrin, Uri Lerner, Ron Parr, Eran Segal, Simon Tong. In 2001, Lise Getoor joined the Computer Science Department at the University of Maryland, College Park. Network Analysis in the Big Data Age: Mining Graph and Social Streams

Using Term Sense to Improve Language Modeling Approach to Genomic IR

Genomic IR, characterized by its highly specific information need, severe synonym and polysemy pr... more Genomic IR, characterized by its highly specific information need, severe synonym and polysemy problem, long term name and rapid growing lit- erature size, is challenging IR community. In this paper, we are focused on ad- dressing the synonym and polysemy issue under the language modeling frame- work. Unlike the ways translation model and traditional query expansion tech- niques approach to this issue, we incorporate term sense into the basic language model, a more fundamental approach to the synonym and polysemy issue in IR. The sense approach not only maintains the simplicity of language models, but also makes the document ranking efficient and effective. A comparative ex- periment on the TREC 2004 Genomic Track data shows significant improve- ment of retrieval performance after incorporating the term sense into a basic language model. The MAP (mean average precision) is significantly raised from 29.17% (the baseline system) to 36.94%. The performance of the sense approach is also...

Topic Signature Language Models for Ad hoc Retrieval

IEEE Transactions on Knowledge and Data Engineering, 2007

Semantic smoothing, which incorporates synonym and sense information into the language models, is... more Semantic smoothing, which incorporates synonym and sense information into the language models, is effective and potentially significant to improve retrieval performance. The previously implemented semantic smoothing models, such as the translation model which statistically maps document terms to query terms, and a number of other works that have followed have shown good experimental results. However, these models are unable to incorporate contextual information. Thus, the resulting translation might be fairly general and contains mixed topics. To overcome this limitation, we propose a novel context-sensitive semantic smoothing (CSSS) method that decomposes a document into a set of weighted context-sensitive topic signatures and then translate those topic signatures into query terms. The language model with such a context-sensitive semantic smoothing is referred to as the topic signature language model in this paper. In detail, we implement two types of topic signatures depending on whether an ontology exists in the application domain or not. One is concept extracted by an ontology-based approach; the other is multiword phrase extracted by Xtract if there is no ontology available in the application domain. The translation probabilities from each topic signature to individual terms are estimated through the EM algorithm. Document models based on topic signature translation are then derived. The new smoothing method is evaluated on TREC 2004/2005 Genomics Track based on the ontology-based concept, and TREC Ad hoc Track (Disk 1, 2 and 3) using multiword phrases. Both experiments show significant improvements in terms of average precision and overall recall over the two-stage language model (TSLM) as well as the language model with context-insensitive semantic smoothing (CISS). .

Mining and state-space modeling and verification of sub-networks from large-scale biomolecular networks

BMC Bioinformatics, 2007

Background: Biomolecular networks dynamically respond to stimuli and implement cellular function.... more Background: Biomolecular networks dynamically respond to stimuli and implement cellular function. Understanding these dynamic changes is the key challenge for cell biologists. As biomolecular networks grow in size and complexity, the model of a biomolecular network must become more rigorous to keep track of all the components and their interactions. In general this presents the need for computer simulation to manipulate and understand the biomolecular network model. Results: In this paper, we present a novel method to model the regulatory system which executes a cellular function and can be represented as a biomolecular network. Our method consists of two steps. First, a novel scale-free network clustering approach is applied to the large-scale biomolecular network to obtain various sub-networks. Second, a state-space model is generated for the sub-networks and simulated to predict their behavior in the cellular context. The modeling results represent hypotheses that are tested against high-throughput data sets (microarrays and/or genetic screens) for both the natural system and perturbations. Notably, the dynamic modeling component of this method depends on the automated network structure generation of the first component and the sub-network clustering, which are both essential to make the solution tractable. Conclusion: Experimental results on time series gene expression data for the human cell cycle indicate our approach is promising for sub-network mining and simulation from large-scale biomolecular network.

Semantic smoothing of document models for agglomerative clustering

In this paper, we argue that the agglomerative clustering with vector cosine similarity measure p... more