文本搜索排序中构造训练集的一种方法

在文本搜索领域,用自学习排序的方法构建排序模型越来越普遍。排序模型的性能很大程度上依赖训练集。每个训练样本需要人工标注文档与给定查询的相关程度。对于文本搜索而言,查询几乎是无穷的,而人工标注耗时费力,所以选择部分有信息量的查询来标注很有意义。提出一种同时考虑查询的难度、密度和多样性的贪心算法从海量的查询中选择有信息量的查询进行标注。在LETOR和从Web搜索引擎数据库上的实验结果,证明利用本文提出的方法能构造一个规模较小且有效的训练集。     

融合多类特征的Web查询意图识别

识别搜索引擎用户的查询意图在信息检索领域是备受关注的研究内容。文中提出一种融合多类特征识别Web查询意图的方法。将Web查询意图识别作为一个分类问题,并从不同类型的资源包括查询文本、搜索引擎返回内容及Web查询日志中抽取出有效的分类特征。在人工标注的真实Web查询语料上采用文中方法进行查询意图识别实验,实验结果显示文中采用的各类特征对于提高查询意图识别的效果皆有一定帮助,综合使用这些特征进行查询意图识别,88。5%的测试查询获得准确的意图识别结果。     

Multi-task learning to rank for web search

Both the quality and quantity of training data have significant impact on the accuracy of rank functions in web search. With the global search needs, a commercial search engine is required to expand its well tailored service to small countries as well. Due to heterogeneous intrinsic of query intents and search results on different domains (i.e., for different languages and regions), it is difficult for a generic ranking function to satisfy all type of queries. Instead, each domain should use a specific well tailored ranking function. In order to train each ranking function for each domain with a scalable strategy, it is critical to leverage existing training data to enhance the ranking functions of those domains without sufficient training data. In this paper, we present a boosting framework for learning to rank in the multi-task learning context to attack this problem. In particular, we propose to learn non-parametric common structures adaptively from multiple tasks in a stage-wise way. An algorithm is developed to iteratively discover super-features that are effective for all the tasks. The estimation of the regression function for each task is then learned as linear combination of those super-features. We evaluate the accuracy of multi-task learning methods for web search ranking using data from multiple domains from a commercial search engine. Our results demonstrate that multi-task learning methods bring significant relevance improvements over existing baseline method.     

Graph based Ranked Answers for Keyword Graph Structure

Keyword query processing over graph structured data is beneficial across various real world applications. The basic unit, of search and retrieval, in keyword search over graph, is a structure (interconnection of nodes) that connects all the query keywords. This new answering paradigm, in contrast to single web page results given by search engines, brings forth new challenges for ranking. In this paper, we propose a simple but effective Fuzzy set theory based Ranking measure, called FRank. Fuzzy sets acknowledge the contribution of each individual query keyword, discretely, to enumerate node relevance. A novel aggregation operator is defined, to combine the content relevance based fuzzy sets and, compute query dependent edge weights. The final rank, of an answer, is computed by non-monotonic addition of edge weights, as per their relevance to keyword query. FRank evaluates each answer based on the distribution of query keywords and structural connectivity between those keywords. An extensive empirical analysis shows superior performance by our proposed ranking measure as compared to the ranking measures adopted by current approaches in the literature.     

Improving keyword based web image search with visual feature distribution and term expansion

This paper discusses techniques for improving the performance of keyword-based web image queries. Firstly, a web page is segmented into several text blocks based on semantic cohesion. The text blocks which contain web images are taken as the associated texts of corresponding images and TF*IDF model is initially used to index those web images. Then, for each keyword, both relevant web image set and irrelevant web image set are selected according to their TF*IDF values. And visual feature distributions of both positive image and negative image are modeled using Gaussian Mixture Model. An image’s relevance to the keyword with respect to visual feature is thus defined as the ratio of positive distribution density over negative distribution density. We combine the text-based relevance model with visual feature relevance model to improve the performance. Thirdly, a query expansion model is used to improve the performance further. Expansion terms are selected according to their cooccurrences with the query terms in the top-relevant set of the original query. Our experiments show that our approach yield significant improvement over the traditional keyword based query model.     

基于流形排序的查询推荐方法

针对传统查询推荐方法中存在的相关性度量问题和冗余性问题,该文中提出了一种新的基于流形排序的查询推荐方法。该方法利用查询数据内在的全局流形结构来获得查询之间的相关性,可以有效避免传统方法中相关性度量对高维稀疏查询数据处理的不足;同时,该方法通过提升结构上具有代表性的查询来达到减小查询推荐的冗余性。在一个大规模商业搜索引擎查询日志上的实验结果表明:使用流形排序的查询推荐方法要优于传统查询推荐方法和现有的Hitting-time Ranking方法。     

A transduction-based approach to fuzzy clustering, relevance ranking and cluster label generation on web search results

This paper details a modular, self-contained web search results clustering system that enhances search results by (i) performing clustering on lists of web documents returned by queries to search engines, and (ii) ranking the results and labeling the resulting clusters, by using a calculated relevance value as a degree of membership to clusters. In addition, we demonstrate an external evaluation method based on precision for comparing fuzzy clustering techniques, as well as internal measures suitable for working on non-training data. The built-in label generator uses the membership degrees and relevance values to weight the most relevant results more heavily. The membership degrees of documents to fuzzy clusters also facilitate effective detection and removal of overly similar clusters. To achieve this, our transduction-based clustering algorithm (TCA) and its fuzzy counterpart (FTCA) employ a transduction-based relevance model (TRM) to consider local relationships between each web document. Results from testing on five different real-world and synthetic datasets results show favorable results compared to established label-based clustering algorithms Suffix Tree Clustering (STC) and Lingo.     

Improving large-scale search engines with semantic annotations

Traditional search engines have become the most useful tools to search the World Wide Web. Even though they are good for certain search tasks, they may be less effective for others, such as satisfying ambiguous or synonym queries. In this paper, we propose an algorithm that, with the help of Wikipedia and collaborative semantic annotations, improves the quality of web search engines in the ranking of returned results. Our work is supported by (1) the logs generated after query searching, (2) semantic annotations of queries and (3) semantic annotations of web pages. The algorithm makes use of this information to elaborate an appropriate ranking. To validate our approach we have implemented a system that can apply the algorithm to a particular search engine. Evaluation results show that the number of relevant web resources obtained after executing a query with the algorithm is higher than the one obtained without it.     

Retrieving keyworded subgraphs with graph ranking score

Keyword queries have long been popular to search engines and to the information retrieval community and have recently gained momentum for its usage in the expert systems community. The conventional semantics for processing a user query is to find a set of top-k web pages such that each page contains all user keywords. Recently, this semantics has been extended to find a set of cohesively interconnected pages, each of which contains one of the query keywords scattered across these pages. The keyword query having the extended semantics (i.e., more than a list of keywords hyperlinked with each other) is referred to the graph query. In case of the graph query, all the query keywords may not be present on a single Web page. Thus, a set of Web pages with the corresponding hyperlinks need to be presented as the search result. The existing search systems reveal serious performance problem due to their failure to integrate information from multiple connected resources so that an efficient algorithm for keyword query over graph-structured data is proposed. It integrates information from multiple connected nodes of the graph and generates result trees with the occurrence of all the query keywords. We also investigate a ranking measure called graph ranking score (GRS) to evaluate the relevant graph results so that the score can generate a scalar value for keywords as well as for the topology.     

Exploiting location information for Web search

Most Web pages contain location information, which are usually neglected by traditional search engines. Queries combining location and textual terms are called as spatial textual Web queries. Based on the fact that traditional search engines pay little attention in the location information in Web pages, in this paper we study a framework to utilize location information for Web search. The proposed framework consists of an offline stage to extract focused locations for crawled Web pages, as well as an online ranking stage to perform location-aware ranking for search results. The focused locations of a Web page refer to the most appropriate locations associated with the Web page. In the offline stage, we extract the focused locations and keywords from Web pages and map each keyword with specific focused locations, which forms a set of <keyword, location> pairs. In the second online query processing stage, we extract keywords from the query, and computer the ranking scores based on location relevance and the location-constrained scores for each querying keyword. The experiments on various real datasets crawled from nj.gov, BBC and New York Time show that the performance of our algorithm on focused location extraction is superior to previous methods and the proposed ranking algorithm has the best performance w.r.t different spatial textual queries.     

基于用户查询日志的网络搜索主题分析

网络搜索分析在优化搜索引擎方面具有举足轻重的作用,而且对用户个人搜索特性进行分析能够提高搜索引擎的精准度。目前,大多数已有模型（比如点击图模型及其变体）,注重研究用户群体的共同特点。然而,关于如何做到既可以获取用户群体共同特点又可以获取用户个人特点方面的研究却非常少。本文研究了基于个人用户网络搜索分析新问题,即通过研究用户搜索的突发性现象,获取个人用户搜索查询的主题分布情况。提出了两个搜索主题模型,即搜索突发性模型（SBM）和耦合敏感搜索突发性模型（CS-SBM）。SBM假设查询词和URL主题是无关的,CS-SBM假设查询词和URL之间是有主题关联的,得到的主题分布信息存储在偏Dirichlet先验中,采用Beta分布刻画用户搜索的时间特性。实验结果表明,每一个用户的网络搜索轨迹都有多种基于用户的独有特点。同时,在使用大量真实用户查询日志数据情况下,与LDA、DCMLDA、TOT相比,本文提出的模型具有明显的泛化性能优势,并且有效地描绘了用户搜索查询主题在时间上的变化过程。     

Lost in PubMed. Factors influencing the success of medical information retrieval

With the explosion of information available on the Web, finding specific medical information in an efficient way has become a considerable challenge. PubMed/MEDLINE offers an alternative to free-text searching on the web, allowing searchers to do a keyword-based search using Medical Subject Headings. However, finding relevant information within a limited time frame remains a difficult task. The current study is based on an error analysis of data from a retrieval experiment conducted at the nursing departments of two Belgian universities and a British university. We identified the main difficulties in query formulation and relevance judgment and compared the profiles of the best and worst performers in the test.For the analysis, a query collection was built from the queries submitted by our test participants. The queries in this collection are all aimed at finding the same specific information in PubMed, which allowed us to identify what exactly went wrong in the query formulation step. Another crucial aspect for efficient information retrieval is relevance judgment. Differences between potential and actual recall of each query offered indications of the extent to which participants overlooked relevant citations.The test participants were divided into “worst”, “average” and “best” performers based on the number of relevant citations they selected: zero, one or two and three or more, respectively. We tried to find out what the differences in background and in search behavior were between these three groups.     

Learning to rank code examples for code search engines

Source code examples are used by developers to implement unfamiliar tasks by learning from existing solutions. To better support developers in finding existing solutions, code search engines are designed to locate and rank code examples relevant to user’s queries. Essentially, a code search engine provides a ranking schema, which combines a set of ranking features to calculate the relevance between a query and candidate code examples. Consequently, the ranking schema places relevant code examples at the top of the result list. However, it is difficult to determine the configurations of the ranking schemas subjectively. In this paper, we propose a code example search approach that applies a machine learning technique to automatically train a ranking schema. We use the trained ranking schema to rank candidate code examples for new queries at run-time. We evaluate the ranking performance of our approach using a corpus of over 360,000 code snippets crawled from 586 open-source Android projects. The performance evaluation study shows that the learning-to-rank approach can effectively rank code examples, and outperform the existing ranking schemas by about 35.65 % and 48.42 % in terms of normalized discounted cumulative gain (NDCG) and expected reciprocal rank (ERR) measures respectively.     

基于用户搜索行为的query-doc关联挖掘

query和doc之间的关联关系是搜索引擎期望获取的一类有价值的信息. query和doc间准确的关联分析不仅可以帮助搜索结果排序,也在query和doc之间的桥接中起到重要作用,以实现相关query和doc之间的信息传递,有利于更深入的query理解和doc理解,并在此基础上开展相关应用.本文提出了一种基于用户搜索行为的query和doc关联关系挖掘算法,该方法首先对用户搜索点击日志中的数据进行整理与分析,构建query与doc间的二部图,再通过采用马尔可夫随机游走模型对二部图数据进行建模,挖掘二部图中的点击数据和session数据,最终挖掘出点击日志中用户没有点击到的doc数据,从而预测出query和doc间的隐含关联关系,同时也可以利用该算法得到query和query潜在的关联关系.基于以上理论基础,我们实现了一套完整的日志挖掘系统,通过大量的实验对比,该系统在各方面均取得了优异的表现,其中对检索结果相关性的性能提升可以达到71.23%,这充分表明,本文所提出的理论和算法能够很好地解决query和doc之间的隐含关系挖掘问题,为提高搜索结果的召回率、实现查询推荐和检索结果聚类奠定了良好的前提基础.     

Query processing over incomplete autonomous databases: query rewriting using learned data dependencies

Incompleteness due to missing attribute values (aka “null values”) is very common in autonomous web databases, on which user accesses are usually supported through mediators. Traditional query processing techniques that focus on the strict soundness of answer tuples often ignore tuples with critical missing attributes, even if they wind up being relevant to a user query. Ideally we would like the mediator to retrieve such possibleanswers and gauge their relevance by accessing their likelihood of being pertinent answers to the query. The autonomous nature of web databases poses several challenges in realizing this objective. Such challenges include the restricted access privileges imposed on the data, the limited support for query patterns, and the bounded pool of database and network resources in the web environment. We introduce a novel query rewriting and optimization framework QPIAD that tackles these challenges. Our technique involves reformulating the user query based on mined correlations among the database attributes. The reformulated queries are aimed at retrieving the relevant possibleanswers in addition to the certain answers. QPIAD is able to gauge the relevance of such queries allowing tradeoffs in reducing the costs of database query processing and answer transmission. To support this framework, we develop methods for mining attribute correlations (in terms of Approximate Functional Dependencies), value distributions (in the form of Naïve Bayes Classifiers), and selectivity estimates. We present empirical studies to demonstrate that our approach is able to effectively retrieve relevant possibleanswers with high precision, high recall, and manageable cost.     

Top-k-size keyword search on tree structured data

Keyword search is the most popular technique for querying large tree-structured datasets, often of unknown structure, in the web. Recent keyword search approaches return lowest common ancestors (LCAs) of the keyword matches ranked with respect to their relevance to the keyword query. A major challenge of a ranking approach is the efficiency of its algorithms as the number of keywords and the size and complexity of the data increase. To face this challenge most of the known approaches restrict their ranking to a subset of the LCAs (e.g., SLCAs, ELCAs), missing relevant results.In this work, we design novel top-k-size stack-based algorithms on tree-structured data. Our algorithms implement ranking semantics for keyword queries which is based on the concept of LCA size. Similar to metric selection in information retrieval, LCA size reflects the proximity of keyword matches in the data tree. This semantics does not rank a predefined subset of LCAs and through a layered presentation of results, it demonstrates improved effectiveness compared to previous relevant approaches. To address performance challenges our algorithms exploit a lattice of the partitions of the keyword set, which empowers a linear time performance. This result is obtained without the support of auxiliary precomputed data structures. An extensive experimental study on various and large datasets confirms the theoretical analysis. The results show that, in contrast to other approaches, our algorithms scale smoothly when the size of the dataset and the number of keywords increase.     

Web search engine: Characteristics of user behaviors and their implication

In this paper, first studied are the distribution characteristics of user behaviors based on log data from a massive web search engine. Analysis shows that stochastic distribution of user queries accords with the characteristics of power-law function and exhibits strong similarity, and the user' s queries and clicked URLs present dramatic locality, which implies that query cache and 'hot click' cache can be employed to improve system performance. Then three typical cache replacement policies are compared, including LRU, FIFO, and LFU with attenuation. In addition, the distribution character-istics of web information are also analyzed, which demonstrates that the link popularity and replica pop-ularity of a URL have positive influence on its importance. Finally, variance between the link popularity and user popularity, and variance between replica popularity and user popularity are analyzed, which give us some important insight that helps us improve the ranking algorithms in a search engine.     

Regularized query classification using search click information

Hundreds of millions of users each day submit queries to the Web search engine. The user queries are typically very short which makes query understanding a challenging problem. In this paper, we propose a novel approach for query representation and classification. By submitting the query to a web search engine, the query can be represented as a set of terms found on the web pages returned by search engine. In this way, each query can be considered as a point in high-dimensional space and standard classification algorithms such as regression can be applied. However, traditional regression is too flexible in situations with large numbers of highly correlated predictor variables. It may suffer from the overfitting problem. By using search click information, the semantic relationship between queries can be incorporated into the learning system as a regularizer. Specifically, from all the functions which minimize the empirical loss on the labeled queries, we select the one which best preserves the semantic relationship between queries. We present experimental evidence suggesting that the regularized regression algorithm is able to use search click information effectively for query classification.     

A survey on search results diversification techniques

The quantity of information placed on the web has been greater than before and is increasing rapidly day by day. Searching through the huge amount of data and finding the most relevant and useful result set involves searching, ranking, and presenting the results. Most of the users probe into the top few results and neglect the rest. In order to increase user’s satisfaction, the presented result set should not only be relevant to the search topic, but should also present a variety of perspectives, that is, the results should be different from one another. The effectiveness of web search and the satisfaction of users can be enhanced through providing various results of a search query in a certain order of relevance and concern. The technique used to avoid presenting similar, though relevant, results to the user is known as a diversification of search results. This article presents a survey of the approaches used for search result diversification. To this end, this article not only provides a technical survey of existing diversification techniques, but also presents a taxonomy of diversification algorithms with respect to the types of search queries.