Attention-based Hierarchical Neural Query Suggestion

Personalized query suggestion based on user behavior

Search systems in online content platforms are typically biased toward a minority of highly consumed items, reflecting the most common user behavior of navigating toward content that is already familiar and popular. Query suggestions are a powerful tool to support query formulation and to encourage exploratory search and content discovery. However, classic approaches for query suggestions typically rely either on semantic similarity, which lacks diversity and does not reflect user searching behavior, or on a collaborative similarity measure mined from search logs, which suffers from data sparsity and is biased by highly popular queries. In this work, we argue that the task of query suggestion can be modelled as a link prediction task on a heterogeneous graph including queries and documents, enabling Graph Learning methods to effectively generate query suggestions encompassing both semantic and collaborative information. We perform an offline evaluation on an internal Spotify dataset of search logs and on two public datasets, showing that node2vec leads to an accurate and diversified set of results, especially on the large scale real-world data. We then describe the implementation in an instant search scenario and discuss a set of additional challenges tied to the specific production environment. Finally, we report the results of a large scale A/B test involving millions of users and prove that node2vec query suggestions lead to an increase in online metrics such as coverage (+1.42% shown search results pages with suggestions) and engagement (+1.21% clicks), with a specifically notable boost in the number of clicks on exploratory search queries (+9.37%).

In this paper we propose a query suggestion method for price comparison search engines. Query suggestion techniques are used for generating alternative queries to facilitate web users in information seeking; in this specific domain, suggestions provided to web users need to be properly generated taking into account that the suggested products must be still available for sale. We propose a novel approach based on a slightly variant of classical query-URL graphs: the query-product click-through bipartite graph. Information extracted both from search engine logs and specific domain features are exploited to build the graph, and one of the advantages of this model is that such a graph can be used to suggest not only related queries but also related products. Concepts used in the proposed method are not restricted to our context but are used in many other major e-commerce and search engine websites, we tested the model on several challenging datasets, and also compared with a recent query suggestion approach specifically designed for price comparison engines. Our solution outperforms the competing approach, achieving higher results in terms of relevance of the provided suggestions and coverage rates on top-8 suggestions.

Query suggestion or auto-completion mechanisms are widely used by search engines and are increasingly attracting interest from the research community. However, the lack of commonly accepted evaluation methodology and metrics means that it is not possible to compare results and approaches from the literature. Moreover, often the metrics used to evaluate query suggestions tend to be an adaptation from other domains without a proper justification. Hence, it is not necessarily clear if the improvements reported in the literature would result in an actual improvement in the users' experience. Inspired by the cascade user models and state-of-the-art evaluation metrics in the web search domain, we address the query suggestion evaluation, by first studying the users behaviour from a search engine's query log and thereby deriving a new family of user models describing the users interaction with a query suggestion mechanism. Next, assuming a query log-based evaluation approach, we propose two new metrics to evaluate query suggestions, pSaved and eSaved. Both metrics are parameterised by a user model. pSaved is defined as the probability of using the query suggestions while submitting a query. eSaved equates to the expected relative amount of effort (keypresses) a user can avoid due to the deployed query suggestion mechanism. Finally, we experiment with both metrics using four user model instantiations as well as metrics previously used in the literature on a dataset of 6.1M sessions. Our results demonstrate that pSaved and eSaved show the best alignment with the users satisfaction amongst the considered metrics.

Query suggestion attracts great concern recently. It is crucial for capturing frequently asked questions in question-answering system and most popular topics in search engine. Besides these, is also used in advertising retrieval systems, e-commerce system for advertising push to get more profits. The paper gives a general review of query suggestion methods. On the whole, all the methods can be grouped into two categories: session based methods and click-through based methods. Adjacency based query suggestion, co-occurrence based query suggestion, query-flow graph based query suggestion, clustering based query suggestion, and bipartite graph based query suggestion are presented respectively in detail. Furthermore, how to evaluate the performance of query suggestion is denoted. Finally the important related issues of the area in further research are discussed.

Search has become a ubiquitous, everyday activity, but finding the right information at the right time in an electronic document collection can still be a very challenging process. Significant time is being spent on identifying suitable search terms, exploring matching documents, rephrasing the search request and assessing whether a document contains the information sought. Once another user is faced with a similar information need, the whole process starts again. There is significant potential in cutting down on this activity by taking a user straight to the required information. As well as delivering technical information and vital regulatory information, a knowledge management solution is concerned with capturing valuable insight and experience in order to share it amongst workers. A search engine has been developed and deployed to technical support staff and we were able to assess its impact on mobile workers. The architecture is based on open-source software to satisfy the basic search functionality, such as indexing, search result ranking, faceting and spell checking. The search engine indexes a number of knowledge repositories relevant to the field engineers. On top of that we have developed an adaptive query suggestion mechanism called Sunny Aberdeen. Query suggestions provide an interactive feature that can guide the user through the search process by providing alternative terminology or suggesting ‘best matches’. In our search engine, the query suggestions are generated and adapted over time using state-of-the-art machine learning approaches, which exploit past user interactions with the search engine to derive query suggestions. Apart from continuously updating the suggestions, this framework is also capable of reflecting current search trends as well as forgetting relations that are no longer relevant. Query log analysis of the system running in a real-life context indicates that the system was able to cut down the number of repeat faults and speeds up the decision process for sending out staff to certain jobs.

Chapter 20 - Query Suggestion with Large Scale Data

With the exponential growth of information on the internet, users have been relying on search engines for finding the precise documents. However, user queries are often short. The inherent ambiguity of short queries imposes great challenges for search engines to understand user intent. Query suggestion is one key technique for search engines to augment user queries so that they can better understand user intent. In the past, query suggestions have been relying on either term-frequency--based methods with little semantic understanding of the query, or word-embedding--based methods with little personalization efforts. Here, we present a sequence-to-sequence-model--based query suggestion framework that is capable of modeling structured, personalized features and unstructured query texts naturally. This capability opens up the opportunity to better understand query semantics and user intent at the same time. As the largest professional network, LinkedIn has the advantage of utilizing a rich amount of accurate member profile information to personalize query suggestions. We applied this framework in the LinkedIn production traffic and showed that personalized query suggestions significantly improved member search experience as measured by key business metrics at LinkedIn.

PurposeThe purpose of this study is to generate diversified results for temporally ambiguous queries and the candidate queries are ensured to have a high coverage of subtopics, which are derived from different temporal periods.Design/methodology/approachTwo novel time-aware query suggestion diversification models are developed by integrating semantics and temporality information involved in queries into two state-of-the-art explicit diversification algorithms (i.e. IA-select and xQuaD), respectively, and then specifying the components on which these two models rely on. Most importantly, first explored is how to explicitly determine query subtopics for each unique query from the query log or clicked documents and then modeling the subtopics into query suggestion diversification. The discussion on how to mine temporal intent behind a query from query log is also followed. Finally, to verify the effectiveness of the proposal, experiments on a real-world query log are conducted.FindingsPreliminary experiments demonstrate that the proposed method can significantly outperform the existing state-of-the-art methods in terms of producing the candidate query suggestion for temporally ambiguous queries.Originality/valueThis study reports the first attempt to generate query suggestions indicating diverse interested time points to the temporally ambiguous (input) queries. The research will be useful in enhancing users’ search experience through helping them to formulate accurate queries for their search tasks. In addition, the approaches investigated in the paper are general enough to be used in many domains; that is, experimental information retrieval systems, Web search engines, document archives and digital libraries.

Users try to articulate their complex information needs during search sessions by reformulating their queries. To make this process more effective, search engines provide related queries to help users in specifying the information need in their search process. In this paper, we propose a customized sequence-to-sequence model for session-based query suggestion. In our model, we employ a query-aware attention mechanism to capture the structure of the session context. is enables us to control the scope of the session from which we infer the suggested next query, which helps not only handle the noisy data but also automatically detect session boundaries. Furthermore, we observe that, based on the user query reformulation behavior, within a single session a large portion of query terms is retained from the previously submitted queries and consists of mostly infrequent or unseen terms that are usually not included in the vocabulary. We therefore empower the decoder of our model to access the source words from the session context during decoding by incorporating a copy mechanism. Moreover, we propose evaluation metrics to assess the quality of the generative models for query suggestion. We conduct an extensive set of experiments and analysis. e results suggest that our model outperforms the baselines both in terms of the generating queries and scoring candidate queries for the task of query suggestion.

Search engine users usually strive to reformulate their queries in the search process to gain useful information. It is hard for search engines to understand users’ search intents and return appropriate results if they submit improper or ambiguous queries. Therefore, query reformulation is a bottleneck issue in the usability of search engines. Modern search engines normally provide users with some query suggestions for references. To help users to better learn their information needs, it is of vital importance to investigate users’ reformulation behaviors thoroughly. In this paper, we conduct a detailed investigation of users’ session-level reformulation behavior on a large-scale session dataset and discover some interesting findings that previous work may not notice before: (1) Intent ambiguity may be the direct cause of long sessions rather than the complexity of users’ information needs; (2) Both the added and the deleted terms in a reformulation step can be influenced by the clicked results to a greater extent than the skipped ones; (3) Users’ specification actions are more likely to be inspired by the result snippets or the landing pages, while the generalization behaviors are impacted largely by the result titles. We further discuss some concerns about the existing query suggestion task and give some suggestions on the potential research questions for future work. We hope that this work could provide assistance for the researchers who are interested in the relative domain.

Traditional query suggestion methods mainly leverage click-through information to find related queries as recommendations, without considering the semantic relateness between queries. In addition, few studies use click-through distribution in diversifying query suggestions. To address these issues, we propose a novel and effective framework to generate relevant and diversified query suggestions. We combine query semantics and click-through information together to generate query suggestion candidates which are highly relevant to original query , we use click-through distribution to diversify the candidates. We evaluate our method on a large-scale search log dataset of a commercial engine, experimental results indicate that our framework has significantly improved the relevance and diversity of suggested queries by comparing to four baseline methods.

Major Web Search Engines take as a common practice to provide Suggestions to users in order to enhance their search experience. Such suggestions have normally the form of queries that are, to some extent, similar to the queries already submitted by the same or related users. The final aim of query suggestions is typically to help users to satisfy their information needs more quickly. In this paper we face this problem from a somewhat different perspective, and we propose a new query suggestion model based on Search Shortcuts, that consist in finding and proposing to the user Successful queries that allowed, in the past, several users to satisfy their information needs. This model differs from traditional query suggestion approaches, and allows the evaluation to be performed effectively by exploiting actual user sessions from the Microsoft 2006 RFP dataset. We evaluate several algorithms applied to this problem, both traditional Collaborative Filtering techniques and ad-hoc solutions, and report on preliminary results achieved.

In recent years, with the development of deep learning, machine translation using neural network has gradually become the mainstream method in industry and academia. The existing Chinese-English machine translation models generally adopt the deep neural network architecture based on attention mechanism. However, it is still a challenging problem to model short and long sequences simultaneously. Therefore, a bidirectional LSTM model integrating attention mechanism is proposed. Firstly, by using the word vector as the input data of the translation model, the linguistic symbols used in the translation process are mathematized. Secondly, two attention mechanisms are designed: local attention mechanism and global attention mechanism. The local attention mechanism is mainly used to learn which words or phrases in the input sequence are more important for modeling, while the global attention mechanism is used to learn which layer of expression vector in the input sequence is more critical. Bidirectional LSTM can better fuse the feature information in the input sequence, while bidirectional LSTM with attention mechanism can simultaneously model short and long sequences. The experimental results show that compared with many existing translation models, the bidirectional LSTM model with attention mechanism can effectively improve the quality of machine translation.

This paper proposes an efficient and effective solution to the problem of choosing the queries to suggest to web search engine users in order to help them in rapidly satisfying their information needs. By exploiting a weak function for assessing the similarity between the current query and the knowledge base built from historical users’ sessions, we re-conduct the suggestion generation phase to the processing of a full-text query over an inverted index. The resulting query recommendation technique is very efficient and scalable, and is less affected by the data-sparsity problem than most state-of-the-art proposals. Thus, it is particularly effective in generating suggestions for rare queries occurring in the long tail of the query popularity distribution. The quality of suggestions generated is assessed by evaluating the effectiveness in forecasting the users’ behavior recorded in historical query logs, and on the basis of the results of a reproducible user study conducted on publicly-available, human-assessed data. The experimental evaluation conducted shows that our proposal remarkably outperforms two other state-of-the-art solutions, and that it can generate useful suggestions even for rare and never seen queries.