Abstract
In Web search, a user first comes up with an information need and issues an initial query. Then some retrieved URLs are clicked and other queries are issued if he/she is not satisfied. We advocate that Web search is governed by a hidden semantic space, and each involved element such as query and URL has its projection, i.e., as a vector, in this space. Each of above actions in the search procedure, i.e. issuing queries or clicking URLs, is an interaction result of those elements in the space. In this paper, we aim at uncovering such a semantic space of Web search that uniformly captures the hidden semantics of search queries, URLs and other elements. We propose session2vec and session2vec+ models to learn vectors in the space with search session data, where a search session is regarded as an instantiation of an information need and keeps the interaction information of queries and URLs. Vector learning is done on a large query log from a search engine, and the efficacy of learnt vectors is examined in a few tasks.
This work is substantially supported by a grant from the Research Grant Council of the Hong Kong Special Administrative Region, China (Project Code: CUHK413510).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Existing methods for vector representation learning [2, 15, 16, 20] cannot be readily applied here due to: (1) our training data is a set of sessions and each of them is represented as a graph, while the training data of existing methods is a set of word sequences; (2) a vector capturing users’ information need is incorporated into our learning procedure. Moreover, we intend to learn a space that uniformly embeds elements of different types such as queries and URLs.
- 2.
The number of contextual elements varies, so we calculate the average of contextual vectors.
- 3.
One may notice that both \(P(s_i;\theta )\) and \(P'(s_i;\theta )\) are defined a s probability of \(s_i\) and they may be unequal. In fact, refer to Eqs. 2, 4, and 5, the probability of a session is calculated from element vectors and parameter vectors associated with the Huffman tree. Therefore, it is possible that different types of input vectors, term-based or element-based, output different values. We would not restrict \(P(s_i;\theta )\) = \(P'(s_i;\theta )\) since such constraint will make the model less flexible in learning vectors for different elements. On the other hand, the session vector \(\mathbf {v}(s_i)\), as an intermediary, softly aligns the dimensions of element vectors and term vectors.
- 4.
References
Baroni, M., Dinu, G., Kruszewski, G.: Don’t count, predict! A systematic comparison of context-counting vs. context-predicting semantic vectors. In: ACL, pp. 238–247 (2014)
Bengio, Y., Ducharme, R., Vincent, P., Janvin, C.: A neural probabilistic language model. J. Mach. Learn. Res. 3, 1137–1155 (2003)
Bing, L., Lam, W., Wong, T.L., Jameel, S.: Web query reformulation via joint modeling of latent topic dependency and term context. ACM Trans. Inf. Syst. 33(2), 1–38 (2015)
Craswell, N., Szummer, M.: Random walks on the click graph. In: SIGIR, pp. 239–246 (2007)
Fang, H., Tao, T., Zhai, C.: A formal study of information retrieval heuristics. In: SIGIR, pp. 49–56 (2004)
Gao, J., He, X., Nie, J.Y.: Clickthrough-based translation models for web search: from word models to phrase models. In: CIKM (2010)
Gao, J., Toutanova, K., Yih, W.T.: Clickthrough-based latent semantic models for web search. In: SIGIR, pp. 675–684 (2011)
Grbovic, M., Djuric, N., Radosavljevic, V., Silvestri, F., Bhamidipati, N.: Context- and content-aware embeddings for query rewriting in sponsored search. In: SIGIR 2015 (2015)
Hofmann, T.: Unsupervised learning by probabilistic latent semantic analysis. Mach. Learn. 42(1–2), 177–196 (2001)
Huang, P.S., He, X., Gao, J., Deng, L., Acero, A., Heck, L.: Learning deep structured semantic models for web search using clickthrough data. In: CIKM, pp. 2333–2338 (2013)
Järvelin, K., Kekäläinen, J.: Cumulated gain-based evaluation of IR techniques. ACM Trans. Inf. Syst. 20(4), 422–446 (2002)
Jones, R., Klinkner, K.L.: Beyond the session timeout: automatic hierarchical segmentation of search topics in query logs. In: CIKM 2008, pp. 699–708 (2008)
Le, Q.V., Mikolov, T.: Distributed representations of sentences and documents. In: ICML, pp. 1188–1196 (2014)
Lee, S., Hu, Y.: Joint embedding of query and ad by leveraging implicit feedback. In: EMNLP, pp. 482–491 (2015)
Mikolov, T.: Statistical language models based on neural networks. Ph.D. thesis (2012)
Mikolov, T., Sutskever, I., Chen, K., Corrado, G.S., Dean, J.: Distributed representations of words and phrases and their compositionality. In: NIPS, pp. 3111–3119 (2013)
Pennington, J., Socher, R., Manning, C.D.: Glove: Global vectors for word representation. In: EMNLP, pp. 1532–1543 (2014)
Ren, X., Wang, Y., Yu, X., Yan, J., Chen, Z., Han, J.: Heterogeneous graph-based intent learning with queries, web pages and wikipedia concepts. In: WSDM, pp. 23–32 (2014)
Robertson, S.E., Walker, S., Jones, S., Hancock-Beaulieu, M., Gatford, M.: Okapi at TREC-3. In: TREC, pp. 109–126 (1994)
Schwenk, H.: Continuous space language models. Comput. Speech Lang. 21(3), 492–518 (2007)
Shen, Y., He, X., Gao, J., Deng, L., Mesnil, G.: A latent semantic model with convolutional-pooling structure for information retrieval. In: CIKM, pp. 101–110 (2014)
Singhal, A., Buckley, C., Mitra, M.: Pivoted document length normalization. In: SIGIR, pp. 21–29 (1996)
Socher, R., Bengio, Y., Manning, C.D.: Deep learning for NLP (without magic). In: Tutorial Abstracts of ACL 2012, p. 5 (2012)
Wu, W., Li, H., Xu, J.: Learning query and document similarities from click-through bipartite graph with metadata. In: WSDM, pp. 687–696 (2013)
Yih, W.t., Toutanova, K., Platt, J.C., Meek, C.: Learning discriminative projections for text similarity measures. In: CoNLL, pp. 247–256 (2011)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing AG
About this paper
Cite this paper
Bing, L., Niu, ZY., Lam, W., Wang, H. (2016). Learning a Semantic Space of Web Search via Session Data. In: Ma, S., et al. Information Retrieval Technology. AIRS 2016. Lecture Notes in Computer Science(), vol 9994. Springer, Cham. https://doi.org/10.1007/978-3-319-48051-0_7
Download citation
DOI: https://doi.org/10.1007/978-3-319-48051-0_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-48050-3
Online ISBN: 978-3-319-48051-0
eBook Packages: Computer ScienceComputer Science (R0)