Data mining learning bootstrap through semantic thumbnail analysis

Abstract

The rapid increase of technological innovations in the mobile phone industry induces the research community to develop new and advanced systems to optimize services offered by mobile phones operators (telcos) to maximize their effectiveness and improve their business. Data mining algorithms can run over data produced by mobile phones usage (e.g. image, video, text and logs files) to discover user's preferences and predict the most likely (to be purchased) offer for each individual customer. One of the main challenges is the reduction of the learning time and cost of these automatic tasks. In this paper we discuss an experiment where a commercial offer is composed by a small picture augmented with a short text describing the offer itself. Each customer's purchase is properly logged with all relevant information. Upon arrival of new items we need to learn who the best customers (prospects) for each item are, that is, the ones most likely to be interested in purchasing that specific item. Such learning activity is time consuming and, in our specific case, is not applicable given the large number of new items arriving every day. Basically, given the current customer base we are not able to learn on all new items. Thus, we need somehow to select among those new items to identify the best candidates. We do so by using a joint analysis between visual features and text to estimate how good each new item could be, that is, whether or not is worth to learn on it. Preliminary results show the effectiveness of the proposed approach to improve classical data mining techniques.

Exploiting visual and text features for direct marketing learning in time and space constrained domains

“…In that paper we showed that the combination of such simple color and text-related features performs better than when each type of feature is considered in isolation. Differently from [8] we propose here to capture the visual content of each offer's image as a whole entity through a holistic representation involving Textons [9][10][11][12]. Experiments presented in Section 5.3 show that a Textons-based representation achieves better results when compared against the representation used in [8].…”

“…Differently from [8] we propose here to capture the visual content of each offer's image as a whole entity through a holistic representation involving Textons [9][10][11][12]. Experiments presented in Section 5.3 show that a Textons-based representation achieves better results when compared against the representation used in [8]. In this paper the Textons-based representation of visual content for each MMS has been used together with text features for learning purpose.…”

“…For instance, the presence of faces in an image can be inferred by the "skin appearance" in the pixel domain; likewise, an outdoor context can be inferred by "sky" and/or "vegetation" appearance [14]. Specifically, the following three types of visual information have been suggested to capture the appearance (see [8] for more details):…”

“…In [8] we proposed some colour based visual features to capture the main appearance of different objects and context. For instance, the presence of faces in an image can be inferred by the "skin appearance" in the pixel domain; likewise, an outdoor context can be inferred by "sky" and/or "vegetation" appearance [14].…”

“…We build on our initial work [8] in which the MMS "goodness" is derived combining text information with color based visual features. In that paper we showed that the combination of such simple color and text-related features performs better than when each type of feature is considered in isolation.…”