Inferring search behaviors using partially observable markov model with duration (POMD)

Yin, He; Wang, Kuansan

doi:10.1145/1935826.1935891

Cited by 14 publications

(17 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The dynamicity in the search session is multidimensional, including users whose behavior might evolve, documents in which content might change, and their context-related relevance with respect to the information need. To tackle this challenge, several approaches aiming at leveraging sequential actions are proposed using Markov Decision models (POMDP) [45,75,139], pattern detection [97], or learning-to-rank methods [50,134].…”

Section: The DI Erent Forms Of Collaborationmentioning

confidence: 99%

On the Collaboration Support in Information Retrieval

Soulier

Tamine

2017

ACM Comput. Surv.

View full text Add to dashboard Cite

Collaborative Information Retrieval (CIR) is a well-known setting in which explicit collaboration occurs among a group of users working together to solve a shared information need. This type of collaboration has been deemed as bene cial for complex or exploratory search tasks. With the multiplicity of factors impacting on the search e ectiveness (e.g., collaborators' interactions or the individual perception of the shared information need), CIR gives rise to several challenges in terms of collaboration support through algorithmic approaches. More particularly, CIR should allow us to satisfy the shared information need by optimizing the collaboration within the search session over all collaborators, while ensuring that both mutually bene cial goals are reached and that the cognitive cost of the collaboration does not impact the search e ectiveness. In this survey, we propose an overview of CIR with a particular focus on the collaboration support through algorithmic approaches. The objective of this article is (a) to organize previous empirical studies analyzing collaborative search with the goal to provide useful design implications for CIR models, (b) to give a picture of the CIR area by distinguishing two main categories of models using the collaboration mediation axis, and (c) to point out potential perspectives in the domain.

show abstract

Section: The DI Erent Forms Of Collaborationmentioning

confidence: 99%

On the Collaboration Support in Information Retrieval

Soulier

Tamine

2017

ACM Comput. Surv.

View full text Add to dashboard Cite

show abstract

“…Implicit feedback consists for instance of query reformulations [71], mouse clicks [95], mouse movements [48,66,67,73,201], measurements of dwell-time [206]-the time users spend on a website-, or even the time a search result [115]. The advantage of implicit feedback over explicit feedback is that it is available in abundance.…”

Section: Interpreting User Interactionsmentioning

confidence: 99%

“…If a user never looked at a search result, their lack of engagement on this result cannot be indicative of low relevance. A number of studies have shown that mouse movement can be an indicator of user examination of search results, and of specific sections within search results [48,66,67,73,201]. Similarly, in a mobile setting recording how long each part of the screen is visible can be considered an indicator of relevance [115].…”

Section: A/b Testingmentioning

confidence: 99%

Search Engines that Learn from Their Users

Schuth

2016

SIGIR Forum

View full text Add to dashboard Cite

More than half the world's population uses web search engines, resulting in over half a billion queries every single day. For many people, web search engines such as Baidu, Bing, Google, and Yandex are among the first resources they go to when a question arises. Moreover, for many search engines have become the most trusted route to information, more so even than traditional media such as newspapers, news websites or news channels on television. What web search engines present people with greatly influences what they believe to be true and consequently it influences their thoughts, opinions, decisions, and the actions they take. With this in mind two things are important, from an information retrieval research perspective. First, it is important to understand how well search engines (rankers) perform and secondly this knowledge should be used to improve them. This thesis is about these two topics: evaluation of search engines and learning search engines. In the first part of this thesis we investigate how user interactions with search engines can be used to evaluate search engines. In particular, we introduce a new online evaluation paradigm called multileaving that extends upon interleaving. With multileaving, many rankers can be compared at once by combining document lists from these rankers into a single result list and attributing user interactions with this list to the rankers. Then we investigate the relation between A/B testing and interleaved comparison methods. Both studies lead to much higher sensitivity of the evaluation methods, meaning that fewer user interactions are required to arrive at reliable conclusions. This has the important implication that fewer users need to be exposed to the results from possibly inferior search engines. In the second part of this thesis we turn to online learning to rank. We learn from the evaluation methods introduced and extended upon in the first part. We learn the parameters of base rankers based on user interactions. Then we use the multileaving methods as feedback in our learning method, leading to much faster convergence than existing methods. Again, the important implication is that fewer users need to be exposed to possibly inferior search engines as they adapt more quickly to changes in user preferences. The last part of this thesis is of a different nature than the earlier two parts. As opposed to the earlier chapters, we no longer study algorithms. Progress in information retrieval research has always been driven by a combination of algorithms, shared resources, and evaluation. In the last part we focus on the latter two. We introduce a new shared resource and a new evaluation paradigm. Firstly, we propose Lerot. Lerot is an online evaluation framework that allows us to simulate users interacting with a search engine. Our implementation has been released as open source software and is currently being used by researchers around the world. Secondly we introduce OpenSearch, a new evaluation paradigm involving real users of real search engines. We describe an implementation of this paradigm that has already been widely adopted by the research community through challenges at CLEF and TREC.

show abstract

“…Mining of click data has become a fertile area for many applications, including result relevance estimation, automatic query suggestion, and many others. One interesting application of using search logs to infer result examination behavior on Search Result Pages (SERP) was introduced by He et al [3]. Another successful application of examination data is identification of relevant sub-parts of documents, which can be used for query expansion or term re-weighting (see Buscher et al [1] for more details).…”

Section: Motivation and Overviewmentioning

confidence: 99%

“…Yet, recent work (e.g., [3]) has been developed based on the assumption that the time a searcher spends examining a particular result abstract or snippet, correlates with result relevance. While this idea is intuitively attractive, to the best of our knowledge it has not been empirically tested.…”

mentioning

confidence: 99%

Re-examining search result snippet examination time for relevance estimation

Lagun

Agichtein

2012

Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

Previous studies of web search result examination have provided valuable insights in understanding and modelling searcher behavior. Yet, recent work (e.g., [3]) has been developed based on the assumption that the time a searcher spends examining a particular result abstract or snippet, correlates with result relevance. While this idea is intuitively attractive, to the best of our knowledge it has not been empirically tested. This poster investigates this hypothesis empirically, in a controlled setting, using eye tracking equipment to compare search result examination time with result relevance. Interestingly, while we replicate previous findings showing examination time to be indicative of whole-page relevance, we find that viewing time of individual results alone is a poor indicator of either absolute result relevance or even of pairwise preferences. Our results should not be taken as negating the usefulness of modeling searcher examination behavior, but rather to emphasize that snippet examination time is not in itself a good indicator of relevance.

show abstract

Inferring search behaviors using partially observable markov model with duration (POMD)

Cited by 14 publications

References 21 publications

On the Collaboration Support in Information Retrieval

On the Collaboration Support in Information Retrieval

Search Engines that Learn from Their Users

Re-examining search result snippet examination time for relevance estimation

Contact Info

Product

Resources

About